Universidade de Aveiro Departamento de Química 2014
Sandra Maria Semedo Carvalho Freire
Caracterização da matéria orgânica solúvel em água de aerossóis urbanos Characterization of water-soluble organic matter from urban aerosols
DOCUMENTO PROVISÓRIO
Universidade de Aveiro Departamento de Química 2014
Sandra Maria Semedo Carvalho Freire
Caracterização da matéria orgânica solúvel em água de aerossóis urbanos Characterization of water-soluble organic matter from urban aerosols Tese apresentada à Universidade de Aveiro para cumprimento dos requisitos necessários à obtenção do grau de Doutor em Química, realizada sob a orientação científica do Doutor Armando da Costa Duarte, Professor Catedrático do Departamento de Química da Universidade de Aveiro, e da Doutora Regina Maria Brandão de Oliveira Duarte, Investigadora Auxiliar do Centro de Estudos do Ambiente e do Mar (CESAM) da Universidade de Aveiro.
Apoio Financeiro do
Instituto Português de Apoio
ao Desenvolvimento (IPAD)
Ao meu filho Rúben,
Aos meus Pais e Irmãos
o júri
presidente Doutor Carlos Manuel Martins da Costa professor catedrático do Departamento de Economia, Gestão e Engenharia Industrial da Universidade de Aveiro
Doutor Armando da Costa Duarte professor catedrático do Departamento de Química da Universidade de Aveiro
Doutora Maria Isabel Almeida Ferra professora catedrática do Departamento de Química da Faculdade de Ciências da Universidade da Beira Interior
Doutora Maria Eduarda da Cunha Pereira professora associada do Departamento de Química da Universidade de Aveiro
Doutora Teresa Alexandra Peixoto da Rocha-Santos professora associada do Instituto Superior de Estudos Interculturais e Transdisciplinares (ISEIT) de Viseu, Instituto Piaget
Doutor Carlos Manuel de Melo Pereira professor auxiliar com agregação do Departamento de Química e Bioquímica da Faculdade de Ciências da Universidade do Porto
agradecimentos Meus sinceros agradecimentos a todas as pessoas que contribuíram para a realização deste trabalho:
Ao Prof. Doutor Armando da Costa Duarte quero manifestar o meu agradecimento pelo incentivo, e pelo empenho demonstrados ao longo do desenvolvimento deste trabalho, agradeço a amizade e compreensão reflectidas na sua preocupação para com o meu bem-estar nos momentos mais delicados deste processo.
À Doutora Regina Maria Brandão de Oliveira Duarte quero expressar a minha admiração pelas suas qualidades científicas, pelo seu profissionalismo, pela seriedade e por todo o tempo dedicado no processo de realização desta tese. O seu entusiasmo por esta pesquisa foi contagiante e deu-me a inspiração para continuar.
Aos meus colegas e amigos de trabalho, Patrícia Santos, Patrícia Silva, Pedro Pato, Cláudia Lopes, Cláudia Mieiro, Pedro Coelho, Ana Teresa, Luciana, António, Bruno, João Matos e Andreia Paula, meus agradecimentos pela agradável convivência e amizade.
Um agradecimento especial à Anabela e à Celine pelo seu companheirismo; pelo acolhimento, pela simpatia e pelo carinho demonstrados ao longo destes anos.
À Universidade de Aveiro, agradeço igualmente o acolhimento e as condições disponibilizadas para a concretização desta tese.
Ao IPAD meus agradecimentos pelo apoio financeiro através da concessão de uma bolsa de Doutoramento, que possibilitou as despesas inerentes ao processo de investigação. Ainda, à Fundação Portuguesa para a Ciência e Tecnologia (FCT) através do projeto ORGANOSOL (FCOMP-01-0124-FEDER-019913; PTDC/CTE-ATM/118551/2010) e ao Centro de Estudos do Ambiente e do Mar (CESAM, Universidade de Aveiro, Portugal), meus agradecimentos pelo apoio financeiro.
Ao Djoy, João, Mário e à Dos Anjos pelo incentivo à realização deste trabalho, e ainda pelo companheirismo demonstrado durante o período deste estudo.
Agradeço a estes e a todos os meus amigos que estiveram sempre presentes e que com as suas palavras de apoio encorajaram-me a continuar.
Aos meus pais e irmãos, meu porto seguro, quero aqui expressar o meu profundo reconhecimento pelo amor e pelo apoio incondicional e por toda a paciência e compreensão nestes anos de ausência.
Ao Ruben, meu filho querido, agradeço a sua existência, agradeço os momentos em que, mesmo à distância a sua meiga voz deu-me mais coragem e vontade de lutar. E peço-te desculpas, meu amor, pela minha ausência em momentos que foram tão especiais para nós.
À Universidade de Cabo Verde, os meus agradecimentos pelas autorizações de dispensa de trabalho concedidas durante a realização do meu Doutoramento.
palavras-chave Aerossóis atmosféricos, área urbana, material carbonáceo, matéria orgânica solúvel em água, composição elementar, caracterização estrutural, espectroscopia, cromatografia líquida bidimensional abrangente.
resumo A matéria orgânica solúvel em água (MOSA) de aerossóis atmosféricos é composta por um conjunto complexo de estruturas moleculares que influenciam as propriedades físico-químicas das partículas atmosféricas e, por conseguinte, desempenham um importante papel em diversos processos atmosféricos globalmente relevantes, afectando o clima e a saúde pública. Devido a uma ampla variedade de fontes e processos de formação, é ainda escasso o conhecimento acerca da composição estrutural da MOSA e do respectivo efeito nas propriedades dos aerossóis atmosféricos.
Assim, esta tese pretende fornecer novas perspetivas sobre a composição molecular da MOSA presente na fracção fina de partículas atmosféricas características de uma área urbana (Aveiro, Portugal). Para o efeito, numa primeira fase do trabalho, foi avaliada a ocorrência de eventuais fenómenos de adsorção de compostos orgânicos semi-voláteis nos filtros de fibra de quartzo utilizados na colheita das amostras de partículas atmosféricas. Posteriormente, e na mesma área urbana, foi efectuada a colheita de amostras de aerossóis atmosféricos, durante um período de 15 meses, numa base de amostragem semanal e em contínuo. Foram efectuados balanços mássicos que permitiram descrever a importância das fracções de carbono elementar, MOSA e matéria orgânica insolúvel em água, na massa total de aerossóis atmosféricos recolhidos na zona urbana de Aveiro, tendo-se dado especial relevo ao estudo dos efeitos de diferentes condições meteorológicas.
Na tentativa de entender a complexidade da MOSA de aerossóis urbanos, foram efectuados estudos de caracterização estrutural com recurso às espectroscopias de infravermelho com transformadas de Fourier (FTIR) acoplada a um sistema de reflectância total atenuada (ATR, sigla inglesa de Attenuated Total Reflectance) (FTIR-ATR) e de ressonância magnética nuclear de
13C com polarização cruzada (CP, sigla inglesa de Cross
Polarization) e rotação em torno do ângulo mágico (MAS, sigla inglesa de Magic Angle Spinning) (RMN CPMAS de
13C) de estado sólido, mas
também através da avaliação da respectiva composição elementar. A caracterização estrutural da MOSA dos aerossóis recolhidos na zona urbana confirmou o carácter heterogéneo deste tipo de matéria orgânica, traduzido por uma multiplicidade de grupos funcionais. De um modo geral, foi possível concluir que as estruturas alifáticas, as estruturas aromáticas, os grupos hidroxilo e os grupos carboxilo constituem funcionalidades comuns às amostras estudadas. A avaliação semi-quantitativa dos dados de RMN CPMAS de
13C mostrou igualmente diferentes distribuições dos
diversos grupos funcionais, entre as amostras de aerossóis colhidos em diferentes períodos sazonais. A presença de sinais típicos de estruturas derivadas de lignina nos espectros de RMN CPMAS de
13C e FTIR-ATR
das amostras de MOSA típicas de estações sazonais mais frias sugere
que as propriedades de MOSA de partículas atmosféricas são influenciadas pelos processos de queima da madeira para aquecimento doméstico.
Complementarmente às técnicas espectroscópicas anteriormente referidas, foi também utilizada a técnica de cromatografia líquida bidimensional abrangente (LC x LC) acoplada aos detectors de fotodíodos, fluorescência e evaporativo com dispersão de luz, com o objectivo de resolver a heterogeneidade das amostras de MOSA e, simultaneamente, mapear a hidrofobicidade versus distribuição de tamanhos moleculares das amostras. A utilização de uma coluna de cromatografia de interacção hidrofílica operada sob condições de fase reversa na primeira dimensão e de uma coluna de cromatografia de exclusão por tamanhos na segunda dimensão, revelou-se muito útil para a separação das amostras de MOSA em frações com hidrofobicidades e tamanhos moleculares distintos. A distribuição de massa molar média (Mw) obtida neste estudo variou entre 48 e 942 Da e 45 a 1241 Da, em termos de detecção por UV e fluorescência, respectivamente. Os resultados obtidos sugerem ainda que as fracções com menor valor de Mw tendem a ter um carácter relativamente mais hidrofóbico.
keywords Atmospheric aerosols, urban area, carbonaceous materials, water-soluble organic matter, elemental composition, structural characterization, spectroscopy, comprehensive two-dimensional liquid chromatography.
abstract Water-soluble organic matter (WSOM) from atmospheric particles comprises a complex array of molecular structures that play an important role on the physic-chemical properties of atmospheric particles and, therefore, are linked to several global-relevant atmospheric processes which impact the climate and public health. Due to the large variety of sources and formation processes, adequate knowledge on WSOM composition and its effects on the properties of atmospheric aerosol are still limited.
Therefore, this thesis aims at providing new insights on the molecular composition of WSOM from fine atmospheric aerosols typical of an urban area (Aveiro, Portugal). In a first step, adsorption phenomena of semi-volatile organic compounds on quartz fibre filters employed in the collection of atmospheric aerosols were assessed. Afterwards, atmospheric aerosol samples were collected during fifteen months, on a weekly basis. A mass balance of aerosol samples was performed in order to set the relative contribution of elemental carbon, WSOM and water-insoluble organic matter to the aerosol mass collected at the urban area of Aveiro, with a special focus on the assessment of the influence of different meteorological conditions.
In order to assess the chemical complexity of the WSOM from urban aerosols, their structural characteristics were studied by means of Fourier transform infrared infrared - Attenuated Total Reflectance (FTIR-ATR) and solid-state cross polarization with magic angle spinning
13C nuclear
magnetic resonance (CPMAS 13
C NMR) spectroscopies, as well as their elemental composition. The structural characterization of aerosol WSOM samples collected in the urban area highlighted a highly complex mixture of functional groups. It was concluded that aliphatic and aromatic structures, hydroxyl groups and carboxyl groups are characteristic to all samples. The semi-quantitative assessment of the CPMAS
13C NMR data showed
different distributions of the various functional groups between the aerosol samples collected at different seasons. Moreover, the presence of signals typical of lignin-derived structures in both CPMAS
13C NMR and FTIR-ATR
spectra of the WSOM samples from the colder seasons, highlights the major contribution of biomass burning processes in domestic fireplaces, during low temperature conditions, into the bulk chemical properties of WSOM from urban aerosols.
A comprehensive two-dimensional liquid chromatography (LC x LC) method, on-line coupled to a diode array, fluorescence, and evaporative light scattering detectors, was employed for resolving the chemical heterogeneity of the aerosol WSOM samples and, simultaneously, to map the hydrophobicity versus the molecular weight distribution of the samples. The LC x LC method employed a mixed-mode hydrophilic interaction
column operating under aqueous reversed phase mode in the first dimension, and a size-exclusion column in the second dimension, which was found to be useful for separating the aerosol WSOM samples into various fractions with distinct molecular weight and hydrophobic features. The estimative of the average molecular weight (Mw) distribution of the urban aerosol WSOM samples ranged from 48 to 942 Da and from 45 to 1241 Da in terms of UV absorption and fluorescence detection, respectively. Findings suggest that smaller Mw group fractions seem to be related to a more hydrophobic nature.
XIII
Table of contents
Table of contents
XIV
Table of contents
XV
Table of contents ............................................................................................................. XIII
List of tables .................................................................................................................... XIX
List of figures ............................................................................................................... XXIII
List of abbreviations .................................................................................................... XXIX
I Objectives & thesis outline ............................................................................................ 1
1.1. Introduction ............................................................................................................ 3 1.2. Objectives of the thesis ........................................................................................... 5
1.3. Structure of the thesis ............................................................................................. 6
II Water-soluble organic matter from atmospheric aerosols: current state of the art 9
2.1. Importance of WSOM in atmospheric aerosols ................................................... 11 2.2. Sources and formation mechanisms of WSOM in atmospheric aerosols............. 12
2.3. Progress and issues in the analysis of WSOM from atmospheric aerosols .......... 15 2.4. Main achievements on the structural characterization of WSOM in atmospheric
aerosols ................................................................................................................. 19 2.5. Molecular weight assessment of WSOM in atmospheric aerosols ...................... 28
III Experimental procedures ............................................................................................ 31
3.1. Introduction .......................................................................................................... 33
3.2. Reagents ............................................................................................................... 33 3.3. Aerosol sampling campaigns ................................................................................ 34
3.4. Determination of organic and elemental carbon content in PM2.5 and PM2.5-10
samples ................................................................................................................. 36
3.5. Extraction of WSOM from the PM2.5 and PM2.5-10 samples ................................ 37 3.6. Determination of dissolved organic carbon content in the aqueous extracts from
the PM2.5 and PM2.5-10 samples ............................................................................. 38 3.7. Isolation and fractionation of WSOM from the PM2.5 samples ........................... 42 3.8. Elemental analysis of the WSOM from the PM2.5 samples .................................. 44
3.9. Spectroscopic characterization of the WSOM from the PM2.5 samples ............... 45
3.9.1. Ultraviolet- Visible (UV-Vis) spectroscopy ............................................ 45
3.9.2. Excitation-Emission matrix (EEM) fluorescence spectroscopy .............. 45
3.9.3. Fourier transform infrared - attenuated total reflectance (FTIR-ATR)
spectroscopy ............................................................................................. 46
3.9.4. Solid-state cross polarization with magic angle spinning 13
C nuclear
magnetic resonance (CPMAS 13
C NMR) spectroscopy ........................... 46
3.10. Comprehensive two-dimensional liquid chromatography of WSOM from PM2.5
samples ................................................................................................................. 47
IV Global carbon balance and isolation of water-soluble organic matter from
atmospheric aerosols .................................................................................................... 49
4.1. Introduction .......................................................................................................... 51
Table of contents
XVI
4.2. Assessment of the meteorological parameters recorded during the sampling
campaigns ............................................................................................................. 51
4.2.1. Meteorological parameters recorded during sampling campaign I .......... 52
4.2.2. Meteorological parameters recorded during sampling campaign II ........ 54
4.3. Assessment of OC adsorption phenomena onto quartz filters during aerosol
sampling ............................................................................................................... 59 4.4. Seasonal trend of fine and coarse aerosol in Aveiro ............................................ 64 4.5. Seasonal sample division and natural event identification ................................... 78 4.6. Impact of forest fire emissions on PM2.5 and carbonaceous material .................. 80 4.7. Isolation and fractionation of water-soluble organic matter from atmospheric
particles ................................................................................................................ 85
4.7.1. Preliminary tests to improve the recovery of organic matter from the
DAX-8 resin ............................................................................................. 85
4.8. Aerosol mass balance of fine atmospheric aerosols ............................................. 91 4.9. Conclusions .......................................................................................................... 94
V Structural characterization of water-soluble organic matter from fine urban
atmospheric aerosols .................................................................................................... 97
5.1. Introduction .......................................................................................................... 99
5.2. UV-Vis and EEM fluorescence spectroscopy of the aerosol water-soluble organic
carbon ................................................................................................................... 99
5.2.1. UV-Vis spectroscopy ............................................................................... 99
5.2.2. EEM fluorescence spectroscopy ............................................................ 104
5.3. Elemental analysis of aerosol water-soluble organic carbon hydrophobic acids 110
5.4. FTIR-ATR spectroscopy of aerosol water-soluble organic carbon hydrophobic
acids .................................................................................................................... 113 5.5. CPMAS
13C NMR spectroscopy of aerosol water-soluble organic carbon
hydrophobic acids ............................................................................................... 116 5.6. Conclusions ........................................................................................................ 121
VI Comprehensive two-dimensional liquid chromatography of water-soluble organic
matter from fine urban atmospheric aerosols ......................................................... 123
6.1. Introduction ........................................................................................................ 125 6.2. Development of PALC × SEC method for the analysis of urban aerosol WSOC
hydrophobic acids ............................................................................................... 128 6.3. Analysis of urban aerosol WSOC hydrophobic acids through PALC × SEC
methodology ....................................................................................................... 132 6.4. Conclusions ........................................................................................................ 143
VII General conclusions ............................................................................................... 145
References......................................................................................................................... 149
Annexes .................................................................................................................................. i
Table of contents
XVII
Annex A ....................................................................................................................... iii
Annex B ........................................................................................................................ v Annex C ....................................................................................................................... ix Annex D .................................................................................................................... xvii
Table of contents
XVIII
XIX
List of tables
List of tables
XX
List of tables
XXI
Table III-1. Results obtained from the procedure adopted in the indirect method in terms of
the analytical signal for TC and IC. ................................................................. 39
Table III-2. Peak area counts acquired through the indirect (analysis of IC followed by TC)
and direct methods using different volumes of sample (ultrapure water). ...... 40
Table III-3. Results obtained from the F test applied to the values of the peak area counts
acquired through the indirect (analysis of IC followed by TC) and direct
methods using ultrapure water. ........................................................................ 40
Table III-4. Effect of the type of HCl container on the average peak area counts (5 replicas,
standard deviation in brackets) obtained in NPOC method using ultrapure
water as a sample. ............................................................................................ 41
Table IV-1. Meteorological data obtained during sampling campaign I (additional
information in section 3.3). .............................................................................. 53
Table IV-2. Average concentration (in µg C m-3
) and standard deviation (in brackets) of
the main carbonaceous components from both PM2.5 and PM2.5-10 in the top
and back filters during sampling campaign I. The number of aerosol samples
in each season was 4. ....................................................................................... 62
Table IV-3. Summarized information regarding the assembled PM2.5 samples collected in
sampling campaign II. ..................................................................................... 79
Table IV-4. Distribution and average percentages (± standard deviation) of retention,
recovery, and loss of UV250 and DOC from three replicas of the
isolation/fractionation procedure of aqueous solutions of Pony Lake fulvic
acids. Three replicas were performed. ............................................................. 85
Table IV-5. Distribution and percentage (average (± sd)) of recovery and loss of UV-Vis
absorbance at 250 nm (UV250) and DOC from the isolation/fractionation
procedure of WSOM from atmospheric samples collected in sampling
campaign II. (1number of replicate=2;
2number of replicate=3) ..................... 87
Table IV-6. Distribution and percentage (average (± sd)) retention and recovery of UV
absorbance at 250 nm (UV250) and DOC from the isolation/fractionation
procedure of WSOM from atmospheric samples collected in the Autumn
season at two different locations. ..................................................................... 88
Table IV-7. Average (Avg.) ambient concentrations (in µg C m-3
) and associated standard
deviation (sd) of total WSOC, and isolated WSOC hydrophobic acids and
WSOC hydrophilic acids fractions. The amount (in mg) of solid residue of
each WSOC hydrophobic acids fraction after freeze-drying is also presented.90
Table V-1. Elemental composition (average (avg) and associated standard deviation (sd))
and atomic ratios of aerosol WSOC hydrophobic acids collected at different
seasons in sampling campaign II. Three replicas were performed. ............... 110
List of tables
XXII
Table V-2. Percentage distribution of carbon in aerosol WSOC hydrophobic acids based
on solid-state CPMAS 13
C-NMR analysis. “N.D.” refers to NMR signal not
detected. ......................................................................................................... 118
Table VI-1. Molecular weight characteristics of the aerosol WSOC hydrophobic acids
from Autumn 2009 and Autumn B 2010, using the PALC × SEC
methodology. ................................................................................................. 138
Table VI-2. Molecular weight characteristics of the aerosol WSOC hydrophobic acids
from Summer B 2010, using the PALC × SEC methodology. ...................... 139
Table VI-3. Molecular weight characteristics of the aerosol WSOC hydrophobic acids
from Winter 2011 and Winter/Spring 2011, using the PALC × SEC
methodology. ................................................................................................. 140
XXIII
List of figures
List of figures
XXIV
List of figures
XXV
Fig. II-1. Pathways for SOA formation in the atmosphere (adapted from Seinfeld and
Pankow (2003)). .............................................................................................. 14
Fig. III-1. Location of the sampling site and picture of the high-volume sampler used in
the aerosol sampling campaigns. ..................................................................... 34
Fig. III-2. Flow diagram for DOC analysis using the indirect (black solid line, TC and IC
measurement) and direct (green dashed line, NPOC measurement) methods. 38
Fig. III-3. Effect of the sparging time in the average peak area counts (grey diamonds, 5
replicas) obtained in NPOC method using ultrapure water as a sample. The
errors bars correspond to the standard deviation. ............................................ 41
Fig. III-4. Example of a calibration graph used for determining the DOC content of the
aerosol aqueous extracts using the NPOC method. ......................................... 42
Fig. III-5. Scheme of the isolation/fractionation procedure of the aerosol WSOM
samples. ........................................................................................................... 44
Fig. IV-1. Weekly and seasonal variability (in terms of median, minimum, and maximum
values) of the (a) air temperature (in ºC), (b) total precipitation accumulated
(in mmH2O), (c) RH (in %), and (d) wind velocity (in ms-1
) recorded during
the sampling campaign II. ............................................................................... 55
Fig. IV-2. Daily variability of the air temperature (in ºC) and HR (in %) recorded
between 1 and 4 January 2010. ........................................................................ 57
Fig. IV-3. Air mass backward trajectories ending at Aveiro at distinct altitudes (>500 m
a.g.l.) during: Autumn 2009 (AVE 1, 23–30 November 2009); Winter 2010
(AVE 10, 2-9 February 2010); Spring 2010 (AVE 20, 27 April to 4 May
2010); Summer 2010 (AVE 35, 09–16 August 2010); Autumn 2010 (AVE 46,
2-9 November 2011), Winter 2011 (AVE 64, 8-15 March 2011), and Spring
2011 (AVE 66, 22-29 March 2011). ................................................................ 58
Fig. IV-4. Distribution of PM2.5 and PM2.5-10 mass concentrations (g m-3
) on the front
and back filters during sampling campaign I. ................................................. 61
Fig. IV-5. PM2.5 and PM2.5-10 concentrations (in μg m-3
) during the annual sampling
campaign II in Aveiro. ..................................................................................... 65
Fig. IV-6. Map of plume of ash trajectories emitted from the Iceland Eyjafjallajökull
volcano eruption in 9th
May 2010, obtained from VAAC at the Meteorological
Office London website
(http://www.metoffice.gov.uk/aviation/vaac/vaacuk_vag.html). .................... 67
Fig. IV-7. Weekly variation of the PM2.5/PM10 ratio during sampling campaign II. ....... 69
Fig. IV-8. Weekly variation of the PM2.5-10 mass concentration (in µg m-3
) and average
concentrations of the carbonaceous fractions (in µg C m-3
) of the PM2.5-10
samples collected during sampling campaign II. Error bars refers to standard
deviation. ......................................................................................................... 71
List of figures
XXVI
Fig. IV-9. Weekly variation of the PM2.5 mass concentration (in µg m-3
) and average
concentrations of the carbonaceous fractions (in µg C m-3
) of the PM2.5
samples collected during sampling campaign II. Error bars refers to standard
deviation. ......................................................................................................... 73
Fig. IV-10. Seasonal distribution of WSOC/OC and OC/EC concentration ratio in PM2.5
fraction during the sampling campaign II. ...................................................... 76
Fig. IV-11. Number of forest fires events recorded during the selected sampling seasons.81
Fig. IV-12. Ambient concentrations of PM2.5 (in µg m-3
) and TC, OC, EC, and WSOC (in
µg C m-3
) for the samples collected during Summer 2008, Summer 2009,
Summer A 2010, and Summer B 2010. ........................................................... 82
Fig. IV-13. WSOC/OC ratio for each PM2.5 sample collected during the warmer periods. 84
Fig. IV-14. Aerosol mass balance for the fine atmospheric particles collected in Aveiro
during Autumn 2009, Winter/Spring 2010, Spring A 2010, Spring B 2010,
Summer A 2010 and Summer B 2010, Autumn A 2010, Autumn B 2010,
Winter 2011, and Winter/Spring 2011 seasons. "NID" refers to the mass of
aerosol that was not identified. ........................................................................ 92
Fig. V-1. UV–Vis spectra (specific absorptivity ɛ (g−1
C L cm−1
) vs. wavelength (nm))
of the WSOC fractions extracted from the aerosol samples collected in the
different seasons. ........................................................................................... 100
Fig. V-2. E2/E3 ratio and ɛ280 (g−1
C L cm−1
) of the aerosol WSOC extracts collected at
the different seasons. ..................................................................................... 103
Fig. V-3. EEM spectra (specific fluorescence intensity (g−1
C L) versus excitation and
emission wavelengths (nm)) of the WSOC fractions extracted from the aerosol
samples collected in Autumn 2009, Spring A 2010, Summer A 2010 and
Winter 2011 seasons. ..................................................................................... 106
Fig. V-4. Synchronous fluorescence spectra with λ of 60 nm of the WSOC extracts
from aerosol samples collected in Autumn 2009, Winter/Spring 2010, Spring
A 2010, Spring B 2010, Summer A 2010, Summer B 2010, Autumn A 2010,
Autumn B 2010, Winter 2011, and Winter/Spring 2011 seasons.................. 108
Fig. V-5. FTIR-ATR spectra (4000–600 cm-1
) of aerosol WSOC hydrophobic acids
samples from the different seasons. ............................................................... 114
Fig. V-6. Solid-state CPMAS 13
C NMR spectra of WSOC hydrophobic acids fractions
of five aerosol samples representative of different seasonal periods. ........... 117
Fig. VI-1. One-dimensional (1D) chromatograms of the urban aerosol WSOC
hydrophobic acids obtained with the PALC technique, and recorded by
different detection methods: (A) DAD operating at 254 nm, FLD at (B)
λExc/λEm = 240/410 nm and (C) λExc/λEm = 320/415 nm, and (D) ELSD.
Additional details about the chromatographic conditions in terms of mobile
List of figures
XXVII
phase composition can be found in section 3.10 (Chapter 3). The flow rate in
PALC was 0.5 ml min−1
. ................................................................................ 129
Fig. VI-2. One-dimensional (1D) chromatograms of the urban aerosol WSOC
hydrophobic acids obtained with the SEC technique, and recorded by different
detection methods: (A) DAD operating at 254 nm, FLD at (B) λExc/λEm =
240/410 nm and (C) λExc/λEm = 320/415 nm, and (D) ELSD. Additional details
about the chromatographic conditions in terms of mobile phase composition
can be found in section 3.10 (Chapter 3). The flow rate in SEC was 2.5 ml
min−1
. ............................................................................................................. 130
Fig. VI-3. PALC × SEC contour plots of aerosol WSOC hydrophobic acids from Autumn
2009 recorded by different detection methods (UV absorption at 254 nm,
fluorescence (FL) at λExc/λEm of 240/410 nm and 320/415 nm, and evaporative
light-scattering (ELSD)). Colours represent the intensity of the analytical
signal. ............................................................................................................. 133
Fig. VI-4. PALC × SEC contour plots of aerosol WSOC hydrophobic acids from
Summer B 2010 recorded by different detection methods (UV absorption at
254 nm, fluorescence (FL) at λExc/λEm of 240/410 nm and 320/415 nm, and
evaporative light-scattering (ELSD)). Colours represent the intensity of the
analytical signal. ............................................................................................ 134
Fig. VI-5. PALC × SEC contour plots of aerosol WSOC hydrophobic acids from Autumn
B 2010 recorded by different detection methods (UV absorption at 254 nm,
fluorescence (FL) at λExc/λEm of 240/410 nm and 320/415 nm, and evaporative
light-scattering (ELSD)). Colours represent the intensity of the analytical
signal. ............................................................................................................. 135
Fig. VI-6. PALC × SEC contour plots of aerosol WSOC hydrophobic acids from Winter
2011 recorded by different detection methods (UV absorption at 254 nm,
fluorescence (FL) at λExc/λEm of 240/410 nm and 320/415 nm, and evaporative
light-scattering (ELSD)). Colours represent the intensity of the analytical
signal. ............................................................................................................. 136
Fig. VI-7. PALC × SEC contour plots of aerosol WSOC hydrophobic acids from
Winter/Spring 2011 recorded by different detection methods (UV absorption
at 254 nm, fluorescence (FL) at λExc/λEm of 240/410 nm and 320/415 nm, and
evaporative light-scattering (ELSD)). Colours represent the intensity of the
analytical signal. ............................................................................................ 137
List of figures
XXVIII
XXIX
List of abbreviations
List of abbreviations
XXX
List of abbreviations
XXXI
ACN Acetonitrile
CCN Cloud condensation nuclei
CPMAS 13
C NMR Cross polarization with magic angle spinning 13
C nuclear
magnetic resonance
DAD Diode array detector
ELSD Evaporative light scattering detector
EC Elemental carbon
EEM Excitation-emission matrix
FLD Fluorescence detector
FTIR-ATR Fourier transformed infrared - attenuated total
reflectance
HILIC Hydrophilic interaction liquid chromatography
IEC Ion-exchange chromatography
LC Liquid chromatography
LC x LC Comprehensive two-dimensional liquid chromatography
MeOH Methanol
Mn Number average molecular weight
MW Molecular weight
Mw Weight average molecular weight
NP Normal phase
NDIR Non-dispersive infrared
NPOC Non-purgeable organic carbon
NMR Nuclear magnetic resonance
OA Organic aerosol
OC Organic carbon
OM Organic matter
PALC Per aqueous liquid chromatography
List of abbreviations
XXXII
PC Pyrolytic carbon
PM2.5 Particulate matter with aerodynamic diameter less than 2.5
μm
PM2.5-10 Particulate matter with aerodynamic diameter between 2.5
and 10 μm
RH Relative humidity
RP Reversed phase
SEC Size-exclusion chromatography
SOA Secondary organic aerosol
SVOC Semi-volatile organic carbon
TC Total carbon
UV-Vis Ultraviolet-Visible
VOC Volatile organic carbon
WINSOC Water-insoluble organic carbon
WSOC Water-soluble organic carbon
WINSOM Water-insoluble organic matter
WSOM Water-soluble organic matter
1st dimension First dimension
2nd
dimension Second dimension
1D-LC One-dimensional liquid chromatography
1
I
Objectives & thesis outline
Chapter I
2
Objectives & thesis outline
3
1.1. Introduction
Atmospheric aerosols were defined by Seinfeld and Pandis (2006) as solid or liquid
particles suspended in the air, comprising a myriad of sizes, ranging from a few
nanometers to several tens of micrometers. Depending on the nature and location of their
sources, these atmospheric particles vary in terms of chemical composition and physical
properties, as well as in terms of spatial and temporal distribution (Poschl, 2005; Seinfeld
and Pandis, 2006). Atmospheric aerosols are typically classified into three categories:
ultrafine mode (aerodynamic diameter (dae) less than 0.1 micrometers, dae < 0.1 m,
designated as PM0.1), fine mode (dae < 2.5 m, designated as PM2.5), and coarse mode (2.5
< dae < 10m, designated as PM2.5-10) (Seinfeld and Pandis, 2006). This division is very
advantageous since fractions of different dae in general have distinct origins, chemical and
physical properties (Jacobson et al., 2000). The nucleation of gas molecules produces
ultrafine particles, which grow rapidly to the fine particle size range by condensation of
precursor gases and by coagulation. These fine airborne particles can further undergo
chemical and physical aging in the atmosphere (e.g., gas uptake or chemical reactions),
thus changing their size, structure, and composition. These ultrafine and fine particles have
both anthropogenic and natural origin and their atmospheric lifetime is days to weeks.
PM2.5-10, on the other hand, is derived mainly from wind-driven or traffic-related
suspension of road, soil, and mineral dust, sea salt, and biological materials (e.g. plant
fragments, microorganisms, pollen). As such, their atmospheric lifetime range from
minutes to hours . The PM2.5 fraction has been related to adverse effects on human health
(namely, some respiratory and cardiovascular diseases) as well as on climate (Poschl,
2005) and, therefore, has been the focus of major scientific and policy concern during the
last two decades.
Atmospheric aerosol particles are composed of a complex mixture of water-soluble
inorganic salts, insoluble mineral dust, and a highly variable fraction of carbonaceous
material having both primary and secondary origin (Seinfeld and Pandis, 2006). This
carbonaceous component is predominantly found in the fine size mode and has been
usually classified as: inorganic carbon (IC), elemental carbon (EC) and organic carbon
(OC) (Seinfeld and Pandis, 2006). The IC fraction typically consists of mineral carbonates
Chapter I
4
and it is derived almost exclusively from soil dust (Seinfeld and Pankow, 2003). The EC
fraction is directly emitted into the atmosphere as a result of incomplete combustion of
fossil fuels and burning biomass, thus having a predominantly primary origin (Seinfeld and
Pandis, 2006). From a chemical and morphological point of view, the EC fraction can be
pictured as more or less disordered stacks of graphene layers or large polycyclic aromatics
with a surface coverage by oxygen-containing functional groups and nitrogen species
(Smith and Chughtai, 1995; Gelencser, 2004). The OC component contributes to 10% to
70% of the total mass of dry fine particles (Turpin et al., 2000) and it is composed of
thousands of organic compounds, ranging from low molecular-weight compounds (e.g.
malonic and oxalic acids) (Sempére and Kawamura, 1994) to n-alkanes, polycyclic
aromatic hydrocarbons, terpens, carbonyls, and n-alkanols (Alves et al., 2002). Several
studies (Facchini et al., 1999; Zappoli et al., 1999; Decesari et al., 2001; Sullivan et al.,
2004; Duarte et al., 2007; Timonen et al., 2008) have further demonstrated that an
important fraction of the OC component in atmospheric aerosols is water-soluble. Studies
performed in both urban (Decesari et al., 2001; Kim et al., 2011) and rural (Duarte et al.,
2007) locations have shown that the water-soluble organic carbon (WSOC) typically
ranged from 35 to 55% of the OC fraction in fine atmospheric aerosols.
WSOC concentrations in fine atmospheric aerosols from urban areas or from places
impacted with biomass burning can reach relatively high values (e.g., 10–15 µg C m-3
(Jaffrezo et al., 2005) or 46 µg C m-3
(Mayol-Bracero et al., 2002), respectively), whereas
in rural areas they can be as low as 5 µg C m-3
(Duarte et al., 2007). In marine
environments, WSOC concentrations can reach up to 0.2 µg C m-3
(Yoon et al., 2007).
This wide range of atmospheric concentration values reflects the relative importance of
primary emissions and/or secondary organic aerosol formation at the different locations.
Duarte et al. (2007) have also shown that the concentrations of WSOC follow a seasonal
trend, with maximum values during Autumn, Autumn/Winter and Winter (colder seasons),
and minimum concentrations during warmer periods. The water-soluble organic matter
(WSOM) comprises a complex mixture likely composed of aliphatic structures,
oxygenated alkyls, carboxylic acids and aromatic structures (Decesari et al., 2000; Duarte
et al., 2005, 2007; Sannigrahi et al., 2006), and contributes up to 10-42% of the total fine
aerosol mass (Duarte, 2006; Sun et al., 2011).
Objectives & thesis outline
5
Due to the large variety of sources and formation processes, adequate knowledge
on WSOM composition and its effects on the properties of atmospheric aerosol are still
limited. The traditional approach of compound speciation by gas and liquid
chromatography coupled to mass spectrometry (GC-MS and LC-MS) provided a long list
of individual organic compounds, which accounts for less than 11% (Graham et al., 2002;
Mayol-Bracero et al., 2002; Decesari et al., 2006) and 14% (Yang et al., 2004) of total
WSOC, respectively. These results suggest that further research is needed focusing on the
detailed characterization of the chemical features (e.g. elemental composition, functional
groups, and fingerprinting) of the bulk WSOM from atmospheric aerosols through a
multidimensional approach based on the synergistic application of advanced analytical
techniques.
1.2. Objectives of the thesis
Taking into account the current state of the art, this PhD project was designed with
the purpose of acquiring new and in-depth knowledge on the carbon chemical
environments of the WSOM from PM2.5 in an urban area. To accomplish this main goal,
aerosol samples collected at Aveiro (Portugal) during fifteen months, between November
2009 and March 2011, will be the basis for the investigation.
The specific objectives of this thesis are:
1. To achieve a global carbon balance (organic carbon (OC) vs. elemental carbon (EC)
vs. water-soluble organic carbon (WSOC)) and a mass closure for the whole mass of
fine aerosol collected, and infer on their seasonal variability;
2. To assess the structural features of WSOM in fine air particles at an urban location,
focusing on their seasonality;
3. To identify fingerprints of possible aerosol emission sources in the integral structural
features of the WSOM;
4. To infer on the contribution of primary and secondary sources to fine particulate OM
through full functional group analysis, and relevant meteorological parameters.
Chapter I
6
It must also be emphasized that atmospheric aerosol sampling employing quartz
fibre filters in high-volume samplers, as performed in this study, are prone to adsorption –
positive bias and overestimation of OC content – or evaporation – negative bias and
underestimation of OC content – of semi-volatile organic carbon (SVOC). As such, the
first task of this research work encompassed the assessment of such phenomena and the
identification of the main limitations of such widely employed methodology.
The structural characterization of the fine urban aerosol WSOM samples was
accomplished by means of UV-Vis, Fourier transformed infrared - attenuated total
reflectance (FTIR-ATR), excitation-emission matrix (EEM) fluorescence, and solid-state
cross-polarization with magic angle spinning 13
C nuclear magnetic resonance (CPMAS 13
C
NMR) spectroscopies. Moreover, a comprehensive two-dimensional liquid
chromatography (LC x LC) technique, employing two independent separation mechanisms
(per-aqueous liquid chromatography (PALC) and size exclusion chromatography (SEC)),
was applied for mapping the hydrophobic character versus molecular weight (MW)
distribution of aerosol WSOM and, simultaneously, offering a new perspective on the
structural heterogeneity of this complex organic aerosol component.
1.3. Structure of the thesis
This thesis is organized into seven chapters, structured as follows:
Chapter I includes a general description of the thesis and a contextualization of the
scientific relevance and specific objectives of the research work.
Chapter II reviews the atmospheric importance of WSOM, its sources and formation
mechanisms, including also the likely trend to achieve a detailed chemical
characterization of WSOM from atmospheric aerosols.
Chapter III presents a description of the materials and experimental methods employed
in this research work, from sample collection to the analytical procedures for structural
characterization of the collected WSOM samples.
Chapter IV describes the assessment of the annual ambient concentrations of PM2.5 and
PM2.5-10, and their TC, OC, EC, and WSOC contents. A global mass balance for the fine
Objectives & thesis outline
7
atmospheric aerosol samples is also presented. This chapter also addresses the
procedure adopted to acquire a representative WSOM sample of the different seasonal
periods, including also the isolation/fractionation procedure employed to obtain a
WSOM fraction amenable to further spectroscopic analysis. This chapter also presents
the results of the SVOC adsorption assessment during the aerosol sampling procedure.
Chapter V presents and discusses the results of the structural characterization of WSOM
from the fine atmospheric aerosols collected in the urban area, focusing on the influence
of different meteorological conditions on their structural composition.
Chapter VI describes the application of LC x LC technique (namely, PALC x SEC) for:
(i) resolving the chemical heterogeneity of the WSOM samples; (ii) determining how
size-distinguished fractions differ in hydrophobicity; and (3) assessing the MW
properties of the studied WSOM samples in regard to number (Mn) and weight (Mw)
average MW, and polydispersity (Mw/Mn).
Chapter VII highlights the main contributions of this research work to the structural
characterization of fine urban aerosol WSOM, and outlines a roadmap for future studies
built upon the work developed in this thesis in order to achieve a detailed structural and
molecular characterization of the complex aerosol WSOM.
Chapter I
8
9
II
Water-soluble organic matter from atmospheric
aerosols: current state of the art
Chapter II
10
Water-soluble organic matter from atmospheric aerosols: current state of the art
11
2.1. Importance of WSOM in atmospheric aerosols
WSOM is a ubiquitous constituent of atmospheric particles and its contribution to
particulate organics in the atmosphere range from 10 to 80% (Duarte and Duarte, 2011a).
Currently, it is recognized the important role played by WSOM on the hygroscopic
properties of aerosols (Dinar et al., 2007), surface- tension behavior (Salma et al., 2006)
and effective density (Dinar et al., 2006) of atmospheric particles and, thus, on the
processes of droplets and cloud formation. As such, WSOM can be an important
contributor to the indirect effects of aerosols over the Earth’s radiative balance (Facchini et
al., 1999; IPCC, 2001; Gysel et al., 2004; Mladenov et al., 2010). Supporting this
assumption some studies can be cited: Gysel et al. (2004) have measured the hygroscopic
properties of the less hydrophilic fraction of WSOM to model the hygroscopic growth and
the authors observed that at 90% relative humidity, 20 to 38 % of particulate water in the
samples is associated to WSOM. Chan et al. (2005) demonstrated that some amines
(namely glycine, alanine, and asparagines) and other organic species (such as
levoglucosan, mannosan, and galactosan) originated from biomass burning could
participate in liquid droplet formation since they can retain water at low relative air
humidity. Using two different aerosol chemical compositions (the classical inorganic case,
and the inorganic plus organic case), Mircea et al. (2002) concluded that the cloud
condensation nuclei (CCN) number concentration depends not only on the percentage of
WSOC, but also on the type of inorganic constituents. Asa-Awuku et al. (2008) also
characterized the CCN activity of water-soluble organics in biomass burning aerosol. The
authors concluded that the presence of salts, primarily (NH4)2SO4, improved substantially
the surface tension decreasing of the most hydrophobic constituents of the WSOC, and
their synergistic effects considerably enhance CNN activity, exceeding that of the pure
inorganic salt. Therefore, increasing the CCN concentrations will result in increased cloud
drop concentrations, smaller drop radii, and more reflective clouds, a phenomenon
generally referred to as the “first indirect effect” (Kanakidou et al., 2005). On the other
hand, a decrease in cloud drop effective radius may lead to lower coalescence rates and,
therefore, reduced precipitation and a longer cloud lifetime, a phenomenon usually referred
to as the “second indirect effect” of atmospheric aerosol (Kanakidou et al., 2005). In either
Chapter II
12
case, it is clear that clouds can significantly impact the amount of solar radiation absorbed
by the planet, and thus affect climate.
Atmospheric aerosols, including their water-soluble and water-insoluble organic
and inorganic fractions, can also influence Earth’s radiative balance directly by scattering
and absorbing the incoming solar radiation. By intercepting incoming solar radiation,
aerosols reduce the energy flux arriving at the Earth’s surface, thus producing a cooling
effect (Charlson et al., 1992). On the other hand, aerosols containing black graphitic and
tarry carbon strongly absorb incoming sunlight. The effect of this type of aerosol are
twofold, both warming the atmosphere and cooling the surface, thus reducing the
atmosphere’s vertical temperature gradient and causing a decline in evaporation and cloud
formation (Kaufman et al., 2002). In the particular case of the WSOM fraction, Hoffer et
al. (2006) investigated the optical properties of WSOM from biomass burning aerosols and
concluded that the relative contribution of WSOM to light absorption in the entire solar
spectrum was of about 7 %. The results of Hoffer et al. (2006) suggest that aerosol WSOM
could have an active, yet highly uncertain, role in the radiative transfer and in
photochemical processes occurring in the atmosphere. To this uncertainty it highly
contributes the lack of adequate knowledge on the chemical composition of aerosol
WSOM, the individual effects of WSOM constituents into the radiative process (absorption
and/or scattering of solar radiation), their emission sources (natural and anthropogenic),
and the synergistic effects resulting from the interactions between the organic and
inorganic fractions and their location (inside or at the surface) in the scattering aerosol
particle.
2.2. Sources and formation mechanisms of WSOM in atmospheric
aerosols
Although the detailed structural characterization of the WSOM in atmospheric
particles is a focus of major scientific concern, its sources and mechanisms of formation is
another emerging issue, for which the atmospheric research community has only
qualitative but not sufficient understanding.
Water-soluble organic matter from atmospheric aerosols: current state of the art
13
Some authors have postulated that WSOM components may be derived primarily
from biomass combustion (Facchini et al., 1999; Zappoli et al., 1999; Mayol-Bracero et al.,
2002; Duarte et al., 2007). Mayol-Bracero et al. (2002) have anticipated different
mechanistic pathways for the formation of WSOM during biomass combustion, including:
(a) soil-derived and/or decaying leaf litter humic matter lofted during combustion, (b)
WSOM generation through chemical transformations of the biomass fuel and/or the initial
volatile combustion products, as well as thermal breakdown of plant lignins and cellulose,
and (c) recombination and condensation reactions of volatile, low molecular weight,
primary products of combustion (Mayol-Bracero et al., 2002). Another potential origin of
WSOM in aerosols includes primary marine sources. Bubble bursting at the ocean surface
has been suggested as a possible mechanistic pathway for the presence of WSOM in
marine aerosols ((Graber and Rudich, 2006) and references therein).
Secondary organic aerosol (SOA) formation may also be a major contributor to the
presence of water-soluble organic species in air particles (Graham et al., 2002; Decesari et
al., 2006; Miyazaki et al., 2006; Duarte et al., 2007; Kondo et al., 2007; Weber et al., 2007;
Saarikoski et al., 2008; Snyder et al., 2009). For example, Saarikoski et al. (2008) reported
for PM1 samples, collected at an urban background site during Summer that, on average,
78% of the SOA is water-soluble. Indeed, SOA formation in the atmosphere has been the
subject of several studies and, according to the critical review of Fuzzi et al. (2006),
different mechanistic pathways for the SOA formation can be put forward: i) partitioning
of semi-volatile organic compounds (SVOC) from the gas phase into/onto pre-existing
particles; ii) participation of SVOC in the formation of new particles (nucleation); and iii)
formation of low- or non-volatile organic compounds by heterogeneous or multiphase
reactions of VOC or SVOC on the surface or in the bulk of aerosol and cloud particle
(atmospheric evolution). The typical SOA formation mechanisms are illustrated in Fig.
II-1.
Chapter II
14
Fig. II-1. Pathways for SOA formation in the atmosphere (adapted from Seinfeld and Pankow (2003)).
The SOA formation by means of a nucleation pathway occurs when VOCs, with
both an anthropogenic and natural origin, are oxidized in the atmosphere to form products
(e.g. polar organic molecules) with lower volatility (SVOCs), which are distributed
between the particulate and the gas phase (Seinfeld and Pankow, 2003). The SVOCs so
formed can condense onto particles already existing in the atmosphere, or they can form
new particles by nucleation. A number of laboratory studies have further suggested the
formation of SOA through photochemical and oxidative processes. For example, SOA
formation from biogenic hydrocarbons through oxidation with hydrogen peroxide under
acidic conditions, and/or through photo-oxidation of biogenic isoprene has been suggested
by Claeys et al. (2004a) and (2004b), respectively. Indeed, isoprene and monoterpenes
emissions from biogenic sources are considered to be the mainly contributors to SOA
formation due to their high emission rates from forest vegetations (annual global emission
of about 500 Tg) (Claeys et al., 2004a). Limbeck et al. (2003) also presented evidences of
SOA formation of atmospheric polymers by means of heterogeneous reaction of dienes
(e.g. isoprene) in the presence of sulfuric acid. Competing oxidants such as ozone or the
presence of humidity decreased the reaction yield, but the formation of polymeric organic
matter was not disabled. Kleindienst et al. (2007) has also suggested that the photo-
oxidation of isoprene, -pinene, -caryophyllene, and toluene are an important source in
the formation of SOA in the atmosphere. Besides biogenic VOCs, the formation of SOA
can also involve VOCs with an anthropogenic origin. On this regard and using WSOC as a
measure of SOA, Weber et al. (2007) provided evidences that SOA was highly correlated
with anthropogenic emissions (carbon monoxide) and anthropogenic secondary organic
VOCs (isopropyl nitrate) in a region containing a mix of both anthropogenic and biogenic
emissions. These findings are also in agreement with those found by Chen et al. (2010),
Water-soluble organic matter from atmospheric aerosols: current state of the art
15
who studied the sources and percursors of SOA in a populated and highly polluted region
during a wintertime episode. The major SOA sources were solvent use (28% of SOA),
catalyst gasoline engines (25% of SOA), wood smoke (16% of SOA), non-catalyst
gasoline engines (13% of SOA), and other unidentified anthropogenic sources (11% of
SOA). In terms of SOA precursors, long-chain alkanes were predicted to be the largest
SOA contributor, followed by aromatic compounds. The study of Duarte et al. (2008b) also
showed evidences that aliphatic material with long-chain (carbons greater than 3 or 4) and
branched mono- and dicarboxylic acids, carbonyl, and ester structural types present in
WSOM samples are likely to be associated with SOA formation from photo-oxidation of
VOCs emitted from either biogenic or local anthropogenic sources.
2.3. Progress and issues in the analysis of WSOM from atmospheric
aerosols
During the past decade, different off-line methods have been developed to deal with
the complexity of WSOM and to acquire knowledge on the structure and composition of
this aerosol component. Commonly-used off-line methods include those focused on
identification and quantification of individual molecular species, and those based on
functional group analysis and characterization of molecular fragments. Analysis of WSOM
based on the typical compound speciation approach usually elucidates less than 20% of the
mass of aerosol WSOM (Facchini et al., 1999; Graham et al., 2002). These methods
typically employ the extraction of the water-soluble organic species from the sampled fibre
filters, followed by LC-MS (Pól et al., 2006; Stone et al., 2009; Zhang et al., 2010) or GC-
MS (Graham et al., 2002; Carvalho et al., 2003; Yu et al., 2004), ion chromatography (IC)
and high-performance liquid chromatography (HPLC) with diode-array detection (DAD)
(HPLC-DAD) (Yang et al., 2004). Off-line techniques that identify functional group
composition, molecular bonds, and molecular fragments, such as one- and two-
dimensional NMR and FTIR spectroscopies and high-resolution MS (Decesari et al., 2000;
Krivacsy et al., 2001a; Graham et al., 2002; Duarte et al., 2005, 2007, 2008b; Sannigrahi et
al., 2006; Reemtsma, 2009; Schmitt-Kopplin et al., 2010), usually provide a more complete
description of WSOM mass, but less detailed information on the individual species (Duarte
Chapter II
16
and Duarte, 2011a). However, before its structural characteristics can be comprehensively
defined through these advanced spectroscopic techniques, the WSOM must be isolated
from other compounds, especially from inorganic species, since they interfere with the
application of such sophisticated analytical techniques. The isolation procedure to be
employed should allow obtaining a representative sample of the original WSOM sample,
without changing its chemical characteristics. In this sense, several methodologies have
been proposed to isolate/fractionate the aerosol WSOM. An in-depth survey of the
literature shows that solid-phase extraction (e.g. hydrophobic bonded-phase silica sorbents
and polymer-based packing materials) (Varga et al., 2001; Andracchio et al., 2002; Kiss et
al., 2002; Duarte and Duarte, 2005; Duarte et al., 2005, 2007, 2008b; Sullivan and Weber,
2006; Sannigrahi et al., 2006; Wozniak et al., 2008; Reemtsma, 2009; Mazzoleni et al.,
2010; Schmitt-Kopplin et al., 2010), ion-exchange chromatography (IEC) (Havers et al.,
1998; Decesari et al., 2000, 2001, 2005; Fuzzi et al., 2001; Graham et al., 2002; Cavalli et
al., 2006), and size-exclusion chromatography (SEC) (Krivacsy et al., 2000; Andracchio et
al., 2002) have been employed to isolate, and simultaneously fractionate, WSOM from the
aerosol aqueous extracts. Of the available isolation procedures, IEC and solid-phase
extraction are currently the most commonly used methods. Isolation procedures based on
the use of IEC exploit the acidic character of the WSOM and allow fractionating the
WSOM mixture into (a) neutral compounds; b) mono- and di-carboxylic acids; and c)
polycarboxylic acids. This procedure does not require any adjustments of the original
acidity of the aerosol aqueous extracts, it allows separating the original WSOM sample
into three main classes of compounds in only one extraction step, and the recoveries are in
the order of 90% of the total mass of WSOM (Decesari et al., 2001). The major drawbacks
are that inorganic species are not removed from the obtained fractions, and the organic
solutes only elute from the resin with high ionic strength solutions (Graber and Rudich,
2006). In what concerns the use of hydrophobic bonded-phase silica (e.g. C18 and Oasis-
HLB) and polymer-based sorbents (e.g. XAD resins), the leading principal is that these
methods fractionate the WSOM samples at pre-adjusted acidic conditions (pH≈2) into
operationally defined hydrophobic and hydrophilic fractions (Kiss et al., 2002; Duarte and
Duarte, 2005; Sannigrahi et al., 2006; Duarte et al., 2007; Schmitt-Kopplin et al., 2010).
With these sorbents, approximately 23-78% of the most hydrophobic and highly
conjugated compounds of the WSOM can be isolated in pure organic form, i.e., free from
Water-soluble organic matter from atmospheric aerosols: current state of the art
17
inorganic species, which are removed through a desalting step (Kiss et al., 2002; Duarte et
al., 2007). It must be emphasized, however, that the application of functionalized solid
sorbents for WSOM extraction has some disadvantages, namely: (i) the back-elution of the
adsorbed organic solutes is time-consuming; (ii) the solid sorbents must be thoroughly
cleaned in order to prevent contamination of the WSOM sample due to resin bleeding; and
(iii) the irreversible adsorption of organic compounds onto the sorbents surfaces (Duarte et
al., 2007).
A special reference is given here to the use of polymer-based sorbents, which were
employed in this work for the isolation/fractionation of aerosol WSOM (Chapter 3, section
3.8). Non-ionic macropourous XAD-8 and XAD-4 resins, in tandem, were first used by
Duarte and Duarte (2005) to separate rural aerosol WSOC samples into hydrophobic acids
and hydrophilic acids fractions. The XAD-8 eluate, which accounted for 47 to 60% of total
WSOC, is represented by partially acidic compounds with significant hydrophobic
moieties (Duarte and Duarte, 2005; Duarte et al., 2007). The XAD-4 eluate holds the most
hydrophilic and of low molecular size compounds of the original WSOC sample, and
accounts for 7 to 12% of the total WSOC (Duarte and Duarte, 2005; Duarte et al., 2007).
Using a method similar to those developed to separate humic and fulvic acids in aqueous
samples, Sullivan and Weber (2006) and Sannigrahi et al. (2006) also applied an XAD-8
resin to separate urban aerosol WSOC into hydrophilic and hydrophobic fractions. The
research paper of Sannigrahi et al. (2006) reports recoveries of about 23% of total WSOC
for the hydrophobic fraction (i.e., from the XAD-8 resin). The major difference between
the methods of Duarte and Duarte (2005) and Sannigrahi et al. (2006) is the solution used
to back elute the organic matter retained in the XAD-8 resin. Sannigrahi et al. (2006)
employed a solution of 0.1 M NaOH, whereas Duarte and Duarte (2005) used a solution of
methanol/ultra-pure water in the proportion 2:3 (40% MeOH). The use of a NaOH solution
has been associated to irreversible hydrolytic reactions of the organic matter. Thus, the
possibility of obtaining a fraction that is not truly representative of the original organic
material is very high and cannot be neglected. In this sense, the use of a 40% MeOH
solution to back elute the retained organic matter is an advantage over the alkaline solution
since it offers the potential to isolate/fractionate the WSOC samples with minimal
introduction of bias due to unwanted reactions (Duarte and Duarte, 2005).
Chapter II
18
The need to study the atmospheric behavior and physico-chemical characteristics of
WSOC on a near real-time scale has also led to the development of methodologies for
continuous on-line measurement of this aerosol component. An approach involving a
particle-into-liquid sampler (PILS) coupled to ion chromatography (IC) was developed in
2001 by Weber and co-workers for a rapid and continuous measurement of the major
inorganic components (e.g. nitrate, sulphate, and ammonium) of aerosol samples (Weber et
al., 2001). This PILS on-line collector was further combined with a total organic carbon
(TOC) analyzer for measurements of total WSOC content in ambient aerosol particles
(Sullivan et al., 2004; Peltier et al., 2007). Aerosol mass spectrometry (AMS) is another
widely used on-line technique that has been used for assessing the chemical submicrometer
non-refractory inorganic and organic particles with high time resolution sensitivity
(DeCarlo et al., 2006). However, AMS does not provide measures of WSOC in
atmospheric aerosols. Furthermore, this technique does not characterize individual
molecules in ambient air. Instead, this on-line technique, together with a custom principal
component analysis, has been used to deconvolute and quantify the mass concentrations of
two types of organic aerosols: oxygenated (OOA) and hydrocarbon-like (HOA) (Zhang et
al., 2005a, 2005b). Kondo et al. (2007) has anticipated that the derived mass concentrations
of OOAs can be used as an indirect measure of the WSOC content in aerosols, assuming
that the organic compounds forming OOAs and WSOCs are similar due to the solubility of
OOAs in water. These authors reported that approximately 88 ± 29% of OOAs is water-
soluble on the basis of the comparison of the WSOC concentrations measured by PILS-
TOC with those of OOA derived from AMS data. Timonen et al. (2010) also reported a
strong correlation (r = 0.88) between aerosol WSOC (measured by PILS-TOC) and
particulate organic matter (measured by AMS), with the WSOC accounting on average for
51% of the particulate organics. Using the AMS mass spectra, Zhang et al. (2005a, 2005b)
also investigated the possible sources of both OOA and HOA. The authors concluded that
OOA is likely SOA (from either anthropogenic or biogenic precursors) and, on a much
lower scale, they have contributions from atmospheric oxidation of HOA and/or biomass
burning. Their observations also suggest that HOA is likely primary aerosol from local,
combustion-related emissions. Overall, it is clear that the on-line AMS is a powerful tool
for routine and real-time measurements of the bulk-chemical nature of OA. However, this
technique has very limited application for the structural characterization of airborne
Water-soluble organic matter from atmospheric aerosols: current state of the art
19
WSOM, being able to provide only a qualitative assessment of the degree of oxidation of
organic aerosols.
2.4. Main achievements on the structural characterization of WSOM in
atmospheric aerosols
The elucidation of the origin and structure of aerosol WSOM have become one of
the main research lines on particulate organics in the atmosphere. Numerous off-line
methodologies have been developed to study the chemical composition of this organic
fraction (Havers et al., 1998; Decesari et al., 2000, 2007; Kiss et al., 2000, 2002; Krivacsy
et al., 2001a; Gelencser et al., 2002; Duarte et al., 2005, 2007, 2008a, 2008b; Sannigrahi et
al., 2006; Reemtsma, 2009; Stone et al., 2009; Schmitt-Kopplin et al., 2010). These
methods are based on a combination of total organic carbon analysis,
isolation/fractionation procedures, and characterization by different analytical techniques
(e.g. NMR, FTIR, UV-Vis and molecular fluorescence spectroscopies, elemental analysis,
and pyrolysis GC-MS).
The elemental analysis has been widely used, in combination with other advanced
analytical techniques, to study natural organic matter (NOM) from different environmental
matrices (e.g. soil, waters, and atmospheric aerosols). Although little structural information
can be drawn from the elemental composition data, the content of C, H, O, and N has been
used to deduce on the origin (terrestrial or microbial) of NOM. Additionally, the atomic
ratio H/C has been also used as an indicator of the amount of saturation of C atoms within
a molecule, whereas the atomic O/C ratio has been assumed to indicate the degree of
oxidation of the NOM samples (Abbt-Braun et al., 2004). The elemental analysis data
reported thus far for aerosol WSOM are in the order of 51-58%, 5.6-6.7%, 2.0-3.8%, and
32-39% for the C, H, N, and O contents, respectively, and regardless of the area under
study (rural areas (Kiss et al., 2002; Duarte et al., 2007) and a high-alpine area (Krivacsy et
al., 2001a)). H/C and O/C values for the aerosol WSOM samples have also been assessed,
being reported atomic ratios in the range of 1.21-1.35 and 0.41-0.55, respectively
(Krivacsy et al., 2001a; Duarte et al., 2007). These data allowed the authors to suggest that
the aerosol WSOM encompass highly oxygenated aliphatic structures. Data on the content
Chapter II
20
of S in aerosol WSOM samples has not been reported yet due to its low concentration,
typically below the detection limit of the analytical instrumentation (Krivacsy et al., 2001a;
Kiss et al., 2002).
The UV-Vis spectroscopy has been quite often applied as a rapid screening method
for the bulk characterization of NOM from different environmental matrices. A typical
UV-Vis spectra of NOM samples, including those of WSOM from aerosols (Krivacsy et
al., 2001a, 2008; Duarte et al., 2003, 2005, 2008a; Baduel et al., 2010) exhibits a
monotonically decrease of the absorbance with increasing wavelength. Although providing
little structural information, the examination of both the quotient E250/E365 (absorbance
values at 250 nm and 365 nm) and specific absorptivity at 280 nm (280) allow some
qualitative information to be withdrawn. According to Peuravuori and Pihlaja (1997), there
is a relationship between the E250/E365 ratio and the percentage of aromaticity and the
molecular size of aquatic humic substances. Higher E250/E365 ratios are usually associated
with lower molecular sizes and lower percentages of aromaticity. On the other hand, higher
280 values have been associated with higher percentages of aromaticity of the NOM
samples (Duarte et al., 2003). Using estimates of these two parameters for assessing the
bulk properties of aerosol WSOC samples collected during Summer and Autumn seasons
at a rural location, Duarte et al. (2005) reported a seasonal trend in the values of the
quotient E250/E365, with higher values being found for the samples collected during
Summer season. Such a pattern allowed the authors to suggest that the WSOM of the
summer samples exhibit a lower molecular size and a lower degree of aromaticity than
those collected in the Autumn season. Duarte et al. (2005) also reported higher 280 values
for the samples collected during the Autumn season, suggesting that these samples include
compounds with complex unsaturated bond systems, where more than two -bond orbital
overlap leading to an increase in the absorptivities values. Krivacsy et al. (2008) also
presented estimates of the quotient E250/E365 for WSOM samples collected in different
locations with different levels of pollution. The authors reported an increase of the
E250/E365 ratio when moving from a polluted urban site to a clean marine environment. The
authors suggested that such a pattern could likely be associated to a decrease in the
aromatic content of the WSOM samples collected at more pristine environments. Krivacsy
et al. (2008) also reported estimates of the specific absorptivity at different wavelengths
(250, 285, 330, 350, and 400 nm). A close examination of the reported values, particularly
Water-soluble organic matter from atmospheric aerosols: current state of the art
21
at 285 nm, shows higher values for samples collected under polluted conditions. This
indicates, as it might be expected, that such samples are likely to be enriched in complex
unsaturated bond systems.
Over the past decades, fluorescence spectroscopy (either emission, excitation,
synchronous, or EEM fluorescence spectroscopy) has been applied extensively as a
relatively simple analytical tool for the characterization, differentiation and classification
of NOM in both terrestrial and aquatic environments (Senesi et al., 1991; Coble, 1996;
Parlanti et al., 2000; McKnight et al., 2001; Santos et al., 2001; Chen et al., 2002;
Peuravuori et al., 2002a). The most efficient fluorophores in NOM are known to be
substituted, condensed aromatic rings, and/or highly unsaturated aliphatic chains
(McDonald et al., 2004). In an opposite way to UV-Vis absorbance, the fluorescence
intensity usually decreases with increasing molecular weight and aromatic content of the
NOM samples (Duarte et al., 2003). The application of molecular fluorescence
spectroscopy, based on the fluorescence emission spectra, in the study of aqueous extracts
of OM aerosols, revealed a single broad band given that the information obtained from this
type of fluorescence spectra is limited. Krivacsy et al. (2001a) and Kiss et al. (2002), have
used the fluorescence spectroscopy to characterize the WSOC fraction in aerosol collected
at high-alpine and at a rural site, as a preliminary evaluation of solid-phase extraction
procedure. The authors obtained a broad band of fluorescence emission with maximum
wavelength at about 405-410 nm when excited with UV wavelength of 235 nm, which is
usually related to polyconjugated molecular structures (e.g., aromatic rings). Duarte et al.
(2004) showed that the EEM fluorescence spectra of WSOC of atmospheric aerosols has
the advantage of obtaining a comprehensive view of the fluorescence characteristics and
allow to identify the major differences in the fluorescent properties of fractions obtained
from the isolation/fractionation procedure of the WSOC of aerosol. In such study, the
authors demonstrated that the fluorescing constituents of WSOM of aerosol from coastal
zone present three fluorescent peaks located at the following excitation/emission (λExc/λEm)
wavelengths: 240/405nm; 310/405nm; and 280/340 nm. The authors also verified that the
first two peaks (λExc/λEm≈240/405 nm and λExc/λEm≈310/405 nm) are located at
wavelengths shorter than those reported for aquatic humic substances, indicating a smaller
content of both aromatic structures and condensed unsaturated bond systems in the WSOC
fraction (Duarte et al., 2004). Duarte et al (2004), also analysed the EEM data to obtain the
Chapter II
22
data of the synchronous spectra at three different λ values (20 nm, 40 nm and 60 nm;
λem=λexc+λ) in order to find the best λ and to found the quickest and most suitable λ for
distinguish samples of WSOC from aerosols with different chemical properties. The
authors concluded that the best peak resolution was achieved by applying a λ of 20 nm,
where two well-defined bands were distinguished at λExc of ≈280 nm and ≈340 nm.
Synchronous spectra profiles of WSOC fractions obtained by the XAD-8/XAD-4 isolation
procedure revealed that the fractions original WSOC and XAD-8 eluate had a shoulder at
λExc ≈ 300 nm, and this shoulder was not present in the synchronous spectra of the other
fractions (XAD-8 effluent, XAD-4 eluate and XAD-4 effluent). The synchronous spectra
of XAD-8 effluent, XAD-4 eluate and XAD-4 effluent presented a peak at lower
wavelength (λExc ≈ 320 nm) than the original WSOC fraction (λExc ≈ 340 nm), suggesting
that such fractions had simpler structural units, with fewer aromatic functional groups and
conjugated systems than the original WSOC sample (Duarte et al., 2004). The synchronous
fluorescence spectra with λ=20 nm was also used by Duarte et al. (2005) for studying the
spectroscopic characteristics of the WSOM isolated from atmospheric aerosols collected at
the rural site under different atmospheric conditions (Summer and Autumn). The authors
observed that all WSOC spectra from atmospheric aerosols exhibited maximum
fluorescence intensities at 285 and 342 nm, and the fluorescence peak at λExc ≈ 285 nm is
usually attributed to aromatic amino acids, phenol-like components and proteinaceous
material, while fluorescence peak at 342 nm is assigned to humic-like compounds (Parlanti
et al., 2000; Yamashita and Tanoue, 2003). Various subsequent studies have also found
protein and humic compounds in WSOC of atmospheric aerosols by fluorescence
spectroscopy (Duarte et al., 2005; Krivacsy et al., 2008; Nakajima et al., 2008; Mladenov
et al., 2010, 2011). For example, Nakajima et al. (2008) studied the fluorescent
characteristics of water-soluble fraction from bulk aerosols collected in urban area of
Okinawa, Japan, based on EEM fluorescence spectra and observed that many water-
soluble samples displayed two peaks at λExc/λEm of 250–275/375–455 and 300–320/400–
440 nm, which can be signatures for fulvic acid-like compounds. The authors also
observed that the several samples collected in Winter periods showed another fluorescence
peak at λExc/λEm of 260–290/305–345 nm, which could be attributed to the presence of
aromatic amines (such as tyrosine and tryptophan).
Water-soluble organic matter from atmospheric aerosols: current state of the art
23
FTIR spectroscopy provides a preliminary assessment of the functional group
composition of WSOM from atmospheric aerosols. The advantages in applying FTIR
spectroscopy to investigate the global structural features of WSOM is that a relatively
small amount of sample is required c.a. 1 mg of WSOM) (Duarte et al., 2007), and the
easiness of obtaining the spectrum. However, the acquired structural information is mainly
qualitative and, in general, complete, unambiguous interpretation of a FTIR spectrum is
rather difficult. A typical FTIR spectrum of WSOM isolated from atmospheric aerosols
exhibits overlapping bands, which clearly indicate multiple functionality as a result of the
complex mixture of organic compounds (Kiss et al., 2002; Duarte et al., 2005, 2007).
Overall, the main functional groups reported in the published studies include aliphatics,
carboxylic acids, hydroxyls, primary amines, carbonyl groups from aldehydes and ketones,
and aromatic structures. The functional groups identified in aerosol WSOM shows
seasonal patterns in their relative importance. For example, Duarte et al. (2005, 2007)
reported the prevalence of aromatic carbons in samples collected in a rural area during
Winter comparatively the Summer samples. The authors attributed this finding to the
presence of lignin breakdown components associated to intensive wood burning in Winter.
Since only some specific bands can be clearly assigned in the FTIR spectrum of such
complex samples, it becomes crucial to corroborate the assignments of the vibrational
frequencies to structural groups using additional spectroscopic methods, such as NMR
spectroscopy.
For appropriate FTIR analysis, some potential interferents such as water soluble
inorganic ions, namely Cl-, K
+, NO3
- and CO3
2- must be previously separated of the organic
fraction. Furthermore, prior to FTIR analysis, the WSOM should be freeze-drieded.
Despite the utility of these procedures, some studies based on FTIR spectroscopy (Maria et
al., 2002; Coury and Dillner, 2009; Hawkins and Russell, 2010; Takahama et al., 2011)
have been applied, by using the direct analysis of the filters substrate onto the equipment.
The results obtained by these two FTIR approaches differ one to the other, the FTIR
spectra obtained by direct filter analysis being less defined because of increasing of
complexity and overlapping. To bypass the problem of complexity and overlapping peaks
by other means without prior removal of interfering species, the FTIR system must be
calibrated using multi-component calibration standards and a multivariate algorithm
(Coury and Dillner, 2008). Although the composition profile is obtained, i.e., several
Chapter II
24
characteristic absorption bands are often seen in the sample, the heterogeneity of aerosol
WSOM produces broad overlapping bands, and the comparison of spectra will only give an
indication of whether the functional groups of each sample are similar or not.
NMR spectroscopy, either one-dimensional liquid-state 1H NMR and solid-state
CPMAS 13
C NMR or two-dimensional liquid-state NMR, has become the most important
spectroscopic method for the structural characterization of complex organic mixtures, such
as those of WSOM from atmospheric aerosols. This spectroscopic technique allows a more
complete description of the whole mass of atmospheric WSOM while providing resolution
on functional groups and sub-structural components. Liquid-state 1H NMR spectroscopy is
the most widely used NMR technique and it has been applied for functional group analysis
(Decesari et al., 2001; Graham et al., 2002; Cavalli et al., 2006), molecular modeling
(Fuzzi et al., 2001), and source apportionment (Decesari et al., 2007) of WSOM from
atmospheric particles. Due to the small chemical shift dispersion of protons (0~10 ppm)
and the chemical complexity of the aerosol WSOM samples, the typical 1H NMR spectrum
consists of a complex overlapping profile, with broad bands superimposed by a relatively
small number of sharp peaks. Usually, only four main categories of functional groups
carrying C-H bonds are identified in the aerosol WSOM samples: 1) Ar-H: aromatic
protons (6.5-8.3 ppm); 2) H-C-O: protons bound to oxygenated aliphatic carbons atoms,
such as aliphatic alcohols, ethers, and esters (3.3-4.1 ppm); 3) H-C-C=: protons bound to
aliphatic carbon atoms adjacent to unsaturated groups, such as alkenes, carbonyl, imino or
aromatic groups (1.9-3.2 ppm); and 4) H-C: aliphatic protons in extended alkyl chains (0.5-
1.9 ppm). Using quantitative integration of each spectral region, it has been concluded that
protons in aliphatic structures are the dominant moieties in atmospheric WSOM, followed
by oxygenated aliphatic compounds and unsaturated aliphatic groups (including H-C-C=O
and H-C-C=C), and only a minor contribution from aromatic groups (Decesari et al., 2001;
Graham et al., 2002; Cavalli et al., 2006). It should be emphasized, however, that 1H NMR
spectroscopy has low sensitivity for detecting functional groups which do not carry protons
(e.g. substituted aromatic compounds), or contain acidic functions with rapidly
exchangeable protons (e.g. carboxylic acids) (Fuzzi et al., 2001; Cook, 2004). In this sense,
solid-state CPMAS 13
C NMR spectroscopy can be viewed as a much more attractive
technique for aerosol WSOM characterization since it gives information about the carbon
backbone of the complex organic structures. However, high quality solid-state CPMAS 13
C
Water-soluble organic matter from atmospheric aerosols: current state of the art
25
NMR spectra are very difficult to obtain due to the low carbon contents of these aerosol
components. In fact, it requires about 20 to 100 mg of sample (depending on the size of the
NMR probe), and the WSOM must also be isolated from the inorganic matrix (including
the naturally occurring paramagnetic trace elements) and prepared as a solid prior to
analysis (Duarte et al., 2005, 2007). Indeed, the only applications to date of CPMAS 13
C
NMR spectroscopy to the study of aerosol WSOM were conducted by Subbalakshmi et al.
(2000), Duarte et al. (2005, 2007) and Sannigrahi et al. (2006). In their studies, the authors
combined the aerosol samples in accordance to similar ambient conditions in order to
obtain sufficient sample mass for the NMR analysis. Despite the above-mentioned
challenges, the CPMAS 13
C NMR technique exhibits important advantages, especially
when compared to one-dimensional liquid-state 1H NMR spectroscopy in the analysis of
aerosol WSOM: a) it is non-destructive and, after analysis, the WSOM samples can be
used for other complementary structural investigations (Duarte et al., 2005, 2007); b) it is
not prone to solvent effects that may alter chemical shifts of the functional groups, mask
some of the chemical resonances due to solvent signals, or even cause the loss of some
peaks (especially those of the carboxylic acids due to the presence of rapidly exchangeable
protons) (Cook, 2004; Sannigrahi et al., 2006); and c) the limited solubility of the organic
material in the selected solvent may result in lower resolution and sensitivity in the liquid-
state 1H NMR spectra (Cook, 2004). Additionally, the CPMAS
13C NMR data can be
employed for obtaining a semi-quantitative measure of the relative contribution of the
different functional groups to the organic carbon present in the aerosol WSOM. These data
can be further used for assessing the structural variations of WSOM with changes in
parameters such as aerosol sources and meteorological conditions (Sannigrahi et al., 2006;
Duarte et al., 2007).
The 13
C chemical shift ranges used to identify WSOM constituents in aerosols are
thoroughly described in the literature (Sannigrahi et al., 2006; Duarte et al., 2007; Duarte
and Duarte, 2009, 2011a), and structural assignment is based on those found for terrestrial
and aquatic natural organic matter (Abbt-Braun et al., 2004; Simpson and Simpson, 2009).
The works reported thus far demonstrate that almost all CPMAS 13
C NMR spectra are very
broad with overlapping peaks, just allowing the identification of typically 5 to 8 types of
functional groups, namely: 0–45 ppm (unsubstituted saturated aliphatic carbons, including
straight-chain methylene (21 ppm) and methane (29 ppm) carbons, and methylene carbons
Chapter II
26
of branched alkyl chains (35 ppm)); 45–60 ppm (substituted aliphatic carbons (e.g., those
found in amines (45 ppm) and methoxyl groups (55 ppm)); 60–95 ppm (oxygenated
aliphatic carbons (e.g., those found in polysaccharides, alcohols, or anhydrosugars)); 95–
110 ppm (aliphatic carbons bonded to two oxygen atoms (e.g., anomeric carbons of
polysaccharides)); 110–140 ppm (alkyl-substituted aromatic carbons and unsaturated
carbons); 140–160 ppm (aromatic carbons bonded to one oxygen atom); 160–190 ppm
(carboxylic, ester and amide carbons); and 190–230 ppm (carbonyl carbons of aldehydes
and ketones) (Duarte and Duarte, 2011a). The published CPMAS 13
C NMR data also
shows evidence that the WSOM from aerosols collected at different locations (rural, urban
and biomass burning) exhibit the same main carbon functional groups, but their relative
abundances are quite different (Sannigrahi et al., 2006; Duarte et al., 2007). Overall, the
results reported thus far show that the WSOM is mostly aliphatic (41–62% of total NMR
peak area), followed by oxygenated aliphatics (15–21%) and carboxylic acid (5.4–13.4%)
functional groups. Duarte et al. (2007) also provided evidences of differences between
aerosol WSOM sources in the warmer and colder periods at a rural location. The aromatic
content of samples collected in Autumn and Winter seasons is higher (15%) than that of
samples collected during the warmer period (6–10%). Furthermore, the Autumn and
Winter samples showed resonances attributable to methoxyl groups (55 ppm) and oxygen-
substituted aromatic ring carbons (147 ppm). These structures were associated with lignin
and its degradation products, which highlights the major contribution of wood-burning
processes in domestic fireplaces to the bulk properties of WSOM from aerosols. The 13
C
NMR results of Sannigrahi et al. (2006) at an urban area also suggested the presence of
aromatic carboxylic acids in the WSOM sample, which were associated with motor-vehicle
emissions or SOA-producing reactions.
Recently, two studies demonstrated the success of two-dimensional liquid-state
NMR techniques in revealing valuable information on the substructures present in WSOM
from atmospheric aerosols (Duarte et al., 2008b; Schmitt-Kopplin et al., 2010). The
combined used of the information provided by 1H-
1H homonuclear (in Correlation
Spectroscopy (COSY) or Total Correlation Spectroscopy (TOCSY)) and 1H-
13C
heteronuclear (in Heteronuclear Single Quantum Coherence (HSQC) or Heteronuclear
Multiple Bond Correlation (HMBC)) connectivities, allows a higher spectral resolution and
therefore greater detail on the C-H backbone of the substructures present in the complex
Water-soluble organic matter from atmospheric aerosols: current state of the art
27
aerosol organic matter mixtures. The HSQC is considered as the most important two-
dimensional NMR method since it provides information on the C-H couplings over one
bond, i.e., it identifies the C-H units in a molecular substructure (Simpson, 2001). After
HSQC, COSY and TOCSY are likely the most useful two-dimensional NMR experiments
for the analysis of organic structures. The COSY method distinguishes protons that are
interacting through one bond, whereas TOCSY identifies protons that are interacting over
two to three bonds (Simpson, 2001). Finally, the HMBC experiment only detects 1H-
13C
couplings over two and three bonds (i.e. H-C-C or H-C-C-C) while single C-H bonds are
absent from the spectrum (Simpson, 2001). This experiment has also an important
advantage over HSQC in that quaternary carbons can be detected with HMBC (Simpson,
2001), thus complementing the chemical shift assignments of the HSQC method. In a
similar fashion of one-dimensional liquid-state 1H NMR spectroscopy, in two-dimensional
NMR experiments it is also important to consider: 1) the amount of sample used (high
sample concentrations could promote molecular aggregation), 2) the sample solubility (the
amount of dissolved organic carbon could be too low for the application of liquid-state 13
C
NMR techniques), and 3) the interference from the solvent signals (Duarte and Duarte,
2013).
In what concerns the structural information provided by these two-dimensional
NMR methods, Duarte et al. (2008b) combined COSY, HSQC, and HMBC experiments
techniques to study WSOM from fine atmospheric aerosols collected during Winter and
Spring/Summer seasons at a rural location with high agricultural activity. The authors
concluded that the aliphatic material of both samples consisted mostly of long-chain
(carbons greater than 3 or 4) and branched mono- and dicarboxylic acids, carbonyl and
ester structural types. The presence of such structural fragments was associated with SOA
formation. Spectral signatures typical of anhydrosugars from cellulose and
methoxyphenols from lignin were also clearly identified among the carbohydrate and
aromatic moieties of the Winter sample. Their presence was linked to the occurrence of
wood burning processes in domestic fireplaces during the colder period. Schmitt-Kopplin
et al. (2010) combined liquid-state 2D NMR techniques (COSY, TOCSY, HSQC, HMBC,
and Distortionless Enhancement by Polarization Transfer (DEPT)-HSQC) with high-
resolution MS to investigate the molecular signatures of the water-soluble fraction of
secondary organic aerosols. The typical aliphatic chemical environment within the studied
Chapter II
28
samples was heteroatom-substituted functional groups adjacent to highly branched
aliphatics (e.g., terpenoid-like molecules). Aromatics were found to be highly substituted,
and electron withdrawing groups and (O)NOx substitution was considerably more common
than the presence of electron donating oxygen-containing functional groups and neutral
substitution (aliphatic carbon). The obtained results allowed the authors to improve the
current knowledge on SOA formation by suggesting possible chemical reaction pathways
involving CHO precursor molecules and sulfuric acid in gas-phase photoreactions.
Despite the serious efforts to unravel the predominant structures in aerosol WSOM,
a comprehensive molecular characterization of this fraction is far from being complete.
Furthermore, linking the composition of WSOM to their sources and formation
mechanisms still remains an issue. Improving the current knowledge of the molecular
structures of aerosol WSOM is therefore highly required to better describe the climate-
relevant properties of atmospheric aerosols.
2.5. Molecular weight assessment of WSOM in atmospheric aerosols
Concerning the assessment of the MW distribution of aerosol WSOM a few studies
have already been reported. Using LC coupled with electrospray ionization mass
spectrometry, Kiss et al. (2003) showed that the number average molecular weight (Mn) of
WSOM collected at rural site is in the range of 215-345 Da. Samburova et al. (2005)
applied laser desorption/ionization mass spectrometry (LDI-MS) to aerosol WSOM
samples collected in an urban ambient. The authors reported a broad range of peaks
between m/z 150 and m/z 500 for the studied samples. Using SEC coupled with diode
array detection, Duarte and Duarte (2011b) reported MW values in the range of 365-1957
Da for a WSOM sample collected during Spring/Summer season, and of 249-1957 Da for a
WSOM sample collected during the Winter season, both at a rural location. Recently,
Wang et al. (2013) characterized the water-soluble organic fraction of carbonaceous
particles, clouds, and fog samples through SEC coupled with inline OC detection. The
authors reported a distribution of WSOM across a wide range of nominal MW (120-10
kDa). The authors also showed that the selected samples presented a substantial fraction of
Water-soluble organic matter from atmospheric aerosols: current state of the art
29
organic matter in a very high MW region (>10 kDa), which the authors associated with
biogenic nanoscale or macromolecular materials.
A recent study conducted by Barros (2011) using LC x LC (specifically, PALC x
SEC) coupled with diode array, fluorescence and evaporative light-scattering (ELSD)
detectors, provided evidences that WSOM from atmospheric particles collected during
Winter season at a rural location has an average MW distribution within the range of 157-
891 Da. This apparent MW distribution of the WSOM sample was found to be lower than
that reported previously by Duarte and Duarte (2011b) using the traditional one-
dimensional SEC approach (i.e., 249-1957 Da). This feature was associated to the
application of a pre-separation procedure (i.e. PALC) in the LC x LC method, which
produces small-size WSOM aggregates. In the conventional SEC procedure reported in the
literature, the WSOM aggregates are analyzed as a whole, thus suggesting that they may
not be so easily dissociated during diffusion through the SEC column, yielding therefore
higher MW values (Duarte et al., 2012). Overall, these results highlight the huge potential
of LC x LC as a promising tool for resolving the chemical heterogeneity of the complex
WSOM in fine atmospheric aerosols. Furthermore, when combined with subsequent offline
structural characterization (e.g., by NMR spectroscopy), LC x LC can be of particular
value for targeting unique molecular structures within the complex aerosol WSOM.
Chapter II
30
31
III
Experimental procedures
Chapter III
32
Experimental procedures
33
3.1. Introduction
This third chapter presents a description of the reagents and experimental methods
employed in this research work, including the sampling procedure for the collection of
PM2.5-10 and PM2.5 samples, the extraction of the water-soluble organic fraction from the
PM2.5 samples, and the procedure for the isolation/fractionation of WSOM from the
aqueous extracts of the collected PM2.5 samples. A description of the analytical procedures
employed in the determination of OC, EC, and WSOC contents in both PM2.5-10 and PM2.5
samples, and in the spectroscopic characterization (UV-Vis, EEM fluorescence, FTIR-
ATR, and CPMAS 13
C NMR spectroscopies) and LC x LC separation of the WSOM from
the PM2.5 samples is also presented.
3.2. Reagents
All the chemicals used in this work were of analytical reagent grade and obtained
from commercial suppliers without further purification. All the solutions were prepared
with high purity water (18 M Ω cm).
The measurement of the dissolved organic carbon (DOC) content in the aerosol
aqueous extracts required the use of hydrochloric acid (HCl) and phosphoric acid (H3PO4)
of ultrapure grade in a glass bottle (purchased from Sigma-Aldrich). A stock DOC standard
solution of 1000 mg C L-1
was monthly prepared from reagent grade potassium hydrogen
phthalate (HO2CC6H4CO2K) (obtained from Merck). This solution was used to prepare,
every week, DOC stock solutions of 100 mg C L-1
. A stock standard solution of 1000 mg C
L-1
of inorganic carbon (IC) was prepared from a mixture of anhydrous sodium carbonate
(Na2CO3) and sodium hydrogen carbonate (NaHCO3) (obtained from Merk). All standard
solutions were stored in dark glass bottles.
Mobile phases for the LC x LC experiments were prepared with HPLC grade
acetonitrile (ACN) obtained from Fisher Scientific (Leicestershire, UK), ammonium
acetate (CH3COONH4), acetic acid (CH3COOH), and ammonium hydrogenocarbonate
(NH4HCO3), all three purchased from Fluka (Buchs, Switzerland). Prior to use, the mobile
Chapter III
34
phases were filtered through hydrophilic polyvinylidene fluoride (PVDF) membrane
(Durapore®, Millipore) of 0.22 m pore size.
3.3. Aerosol sampling campaigns
Atmospheric aerosol samples were collected in the campus of University of Aveiro
(40°38'N, 8°39'W), which is located at about 7 km from the coast on the outskirts of the
small town of Aveiro with approximately 78.400 inhabitants and an urban area of 13.5 km²
(Fig. III-1). The sampling site is placed at 2 m above the ground level (Fig. III-1), about 1
km away from the A25 motorway, and adjacent to a residential area and a salt pans area of
the Lagoon of Ria de Aveiro. An industrial complex, which includes the production of
nitric acid, aniline and nitrobenzene, is located at about 10 km to the north of Aveiro. The
sampling site is impacted by both marine air masses travelling from the Atlantic Ocean and
anthropogenic emissions from vehicular transport, residential, and industrial sources.
Fig. III-1. Location of the sampling site and picture of the high-volume sampler used in the aerosol
sampling campaigns.
Experimental procedures
35
The particulate matter was sampled with a high-volume air sampler, Model TE-
6070V, from Tisch Environmental, Inc. Aiming at separate particles into two size fractions
(dae < 2.5 m and 2.5 < dae < 10 m), the high-volume sampler was equipped with a size
selective PM10 inlet (Model TE-6001, Tisch Environmental, Inc.) and a PM2.5 single-stage
impactor (Model TE-231, Tisch Environmental, Inc.). Sampling was carried out on pre-
fired (at 500 °C for 12h) quartz fibre filters (20.3×25.4 cm; Whatman QM-A) with an air
flow rate of 1.13 m3 min
−1. The aerosols samples were collected on a weekly basis (7 days)
in order to accumulate enough material for the analysis.
In a first stage, and in order to assess positive sampling artifacts in particulate OC
measurements in an urban environment, both PM2.5 and PM2.5-10 samples were collected
using a tandem filter method, i.e., on sets of two (front and back) quartz fibre filters
directly on top of each other. In this field campaign, hereafter designated as sampling
campaign I, a total of 12 filter pairs (Table IV-1, section 4.2.1) were collected from 23
June to 29 July 2008 (Summer 2008), 27 April to 01 June 2009 (Spring 2009), and 01 June
to 09 August 2009 (Summer 2009). In a second stage, and in order to assess the seasonal
trend of the chemical composition of WSOM in fine urban aerosol samples, a second
intensive field campaign, hereafter designated as sampling campaign II, was carried out
from 23 November 2009 to 29 March 2011. The sampling details of the intensive field
campaign II is summarized in Table A. 1, Annex A. In this second field campaign, a total
of 47 aerosol samples (PM2.5 and PM2.5-10) and 11 field blank samples were collected using
the conventional sampling approach, i.e., one filter placed at each size range. The field
blank samples were prepared, stored, transported and analysed in the same way as the
exposed filter samples, with the exception that no air was forced across the filters, in order
to control possible sample contamination, including field operations. During this second
sampling campaign, several aerosol samples could not be collected properly as result of
equipment breakdown associated with bad weather conditions and electricity supply
failures, mainly in January 2010 and 2011, and March 2010, and damaged internal motor
parts (mostly, brushes). The 47 aerosol samples were grouped together (groups of three,
four, and five samples), according to similar ambient conditions, on a total of 10 groups
representative of different seasonal periods (additional information is provided in section
4.5, Chapter 4). The meteorological data, including air temperature, relative humidity
(RH), precipitation, and wind velocity, are summarized in section 4.2 of Chapter 4. These
Chapter III
36
data were provided by the Portuguese Sea and Atmosphere Institute and by the Department
of Physics from the University of Aveiro.
In both sampling campaigns, the concentrations of both PM2.5 and PM2.5-10 (in g
m-3
) were determined by weighting the filters under controlled moisture conditions before
and after exposure. Before sampling, the pre-fired quartz fibre filters were placed inside a
box for 24h at room temperature and approximately 50% relative humidity. The filters
were then weighted in an analytical balance (OHAUS, Model PioneerTM
Analítica-
PA64C), and the weighing precision was ±0.0001 g. After sampling, filter samples were
folded in two, with the exposed side face to face, wrapped in pre-fired aluminium foil and
immediately transported to the laboratory, where they were again placed inside the above-
mentioned box for 24h and weighted. Finally, the field blank and loaded filter samples
were stored frozen until further analysis.
3.4. Determination of organic and elemental carbon content in PM2.5
and PM2.5-10 samples
The determination of OC and EC content in the aerosol samples was performed at
the Portuguese Agency of Environment, using a Lab OC-EC Aerosol Analyzer (Sunset
Laboratory Inc.) and following a thermo-optical method (Birch and Cary, 1996). As with
other procedures (thermal and thermo-optical) applied into the analysis of OC and EC in
aerosol samples, the contents of OC and EC are operationally defined. A brief description
of the procedure follows:
Part A: in a completely oxygen (O2)-free helium (He) atmosphere, a small portion (1.5
cm2) of the quartz fibre filter is heated in four increasing temperature steps (60 sec at
315 ºC, 60 sec at 475 ºC, 60 sec at 615 ºC, and 90 sec at 870ºC) to remove all OC from
the filter. The transition from the third to the fourth temperature (i.e., from 615 ºC to
870 ºC) quickly decomposes inorganic carbonates, producing a sharp, characteristic
peak. During this first stage of analysis, a fraction of OC is usually pyrolitically
converted to EC, called pyrolytic carbon (PC). This pyrolitic conversion is continuously
monitored by measuring the transmission of a laser through the filter. As the organic
Experimental procedures
37
compounds are vaporized, they are immediately oxidized to carbon dioxide (CO2) in an
oxidizer oven which follows the sample oven. The flow of He, containing the CO2, then
goes to a methanator oven where the CO2 is reduced to methane (CH4). The CH4 is then
detected by a flame ionization detector (FID).
Part B: at the end of the fourth temperature (Part A), the sample oven is cooled to 550
ºC (50 sec). After this step, the pure He is switched to a 2% O2/He mixture in the
sample oven and the temperature is stepped up to 940 °C. During this second stage of
analysis, both the original EC and the PC (Part A) are oxidized to CO2 due to the
presence of O2 in the carrier gas. As previously, the CO2 is then reduced to CH4 and
detected by the FID. The darkness of the filter is continuously monitored throughout all
stages of the analysis.
Part C: after all carbon has been oxidized from the sample, a known volume and
concentration of CH4 is injected into the sample oven in order to calibrate each
measurement to a known quantity of carbon. This also provides a means of checking the
operation of the instrument. Based on the FID response and laser transmission data, the
quantities of OC and EC are calculated for each sample. The TC content of the aerosol
sample is determined as OC + EC.
3.5. Extraction of WSOM from the PM2.5 and PM2.5-10 samples
The extraction of WSOC from the PM2.5 and PM2.5-10 samples was based on the
procedure developed by Duarte and Duarte (2005). In this procedure, 280 cm2 and 25 cm
2
of each PM2.5 and PM2.5-10 filter sample, respectively, was extracted with 150 and 30 mL of
ultrapure water, respectively, by mechanical stirring during 5 min plus ultrasonic bath
during 15 min. The final slurry so obtained was filtered through a membrane filter (PVDF,
Durapore®, Millipore) of 0.22 µm pore size.
Chapter III
38
3.6. Determination of dissolved organic carbon content in the aqueous
extracts from the PM2.5 and PM2.5-10 samples
The dissolved organic carbon (DOC) content of each aqueous extract of the aerosol
samples was measured by means of a Shimadzu TOC-5000A analyser with a regular
sensitive catalyst, which offers two independent methods for measuring DOC: i) an
indirect method, where the DOC content is calculated as the difference between the total
carbon (TC) and inorganic carbon (IC) contents; and ii) a direct method, named Non-
Purgeable Organic Carbon (NPOC), where the sample is acidified and sparged with high
purity air for eliminating the IC component prior to DOC measurement. The principle of
DOC analysis is based on the injection of the sample into an oven where it is catalytically
oxidized to CO2 at 680 °C, being then purified by filters, dried, and flushed into the non-
dispersive infrared (NDIR) detector by means of a carrier gas flow. Fig. III-2 schematically
illustrates the indirect and direct methods for DOC analysis.
Fig. III-2. Flow diagram for DOC analysis using the indirect (black solid line, TC and IC measurement)
and direct (green dashed line, NPOC measurement) methods.
Four-port valve
NDIR
Purified air
Sparge gas flowmeter
Sample vessel
Microliter syringe
Cooling tube
Halogen scrubber
TC
IC
NPOC
TC furnace (680º)
Catalyst
IC reaction vessel
Experimental procedures
39
Although the methods used here are standard protocols, the analytical nature of this
work requires their description. A key point about analysing aqueous samples with low
DOC concentrations, such as those of atmospheric aerosol samples, is the need to ensure
the precision of the analytical method (indirect and direct) and, simultaneously, avoid any
possible disturbance/contamination during DOC analysis. A few examples of possible
sources of disturbance/contamination during DOC analysis include the HCl stock bottle (in
direct method), time of sparging in the direct method, and the intensity of the analytical
signal of the blanks (ultra-pure water). Thus, in this work, the results obtained by means of
the indirect and direct methods were compared and the effect of the procedure adopted in
each method was assessed: a) in the indirect method, the analysis of TC followed by the
analysis of IC or the analysis of IC followed by the analysis of TC, and the volume of the
sample injected into the analyser; b) in the direct method, the adequate time of sparging,
the volume of the sample injected into the analyser, and the type of container (glass or
plastic) used to preserve the HCl reagent.
As shown in Table III-1, in the indirect method, the measurement of IC content
previously to that of TC yielded a lower average intensity for the analytical signal acquired
in the analysis of ultrapure water than when performing the opposite sequence.
Table III-1. Results obtained from the procedure adopted in the indirect method in terms of the analytical
signal for TC and IC.
Procedure
TC
(Peak area count) IC
(Peak area count) Volume of sample
(ultrapure water)
Avg a sd
b Avg
a sd
b
TC followed by IC 608 52 550 107 100 µl
IC followed by TC 372 41 370 72
a Avg: average of 5 measurements;
b sd: standard deviation.
When assessing the most appropriate volume of sample (ultrapure water) to be
injected into the analyser, it was found that an injection volume of 25 L yielded the
lowest variability for the analytical signal in both indirect (analysis of IC followed by TC)
and direct methods, as shown in Table III-2.
Chapter III
40
Table III-2. Peak area counts acquired through the indirect (analysis of IC followed by TC) and direct
methods using different volumes of sample (ultrapure water).
Volume of sample
(µL)
Indirect Method a Direct Method
difference of Avg b
(Peak area count) sd
c
(Peak area count) Avg
b
(Peak area count) sd
c
(Peak area count)
10 -221 186 198 28
25 144 116 324 41
50 683 317 571 176 a Value of the analytical signal = Peak area count of TC – Peak area count of IC;
b Avg: average of 10 measurements;
c sd: standard deviation.
As such, it was decided to adopt an injection volume of 25 L for the subsequent
analyses. Using the values of the peak area counts obtained for ultrapure water with this
injection volume, the following step was focused on testing whether the direct method is
more precise than the indirect method through an F-test. As shown in Table III-3, the
calculated value of F exceeds that of the critical value, thus indicating that the variance of
the indirect method is significantly greater than that of the direct method at the 5%
probability level, i.e. the direct method is more precise.
Table III-3. Results obtained from the F test applied to the values of the peak area counts acquired through
the indirect (analysis of IC followed by TC) and direct methods using ultrapure water.
Volume of
sample
Analytical
method
Peak area count Degrees of
freedom Fcalculated
Fcritical
(P = 0.05) Avg a sd
b
25 µL Indirect 144 116
9 8.005 3.179 Direct 324 41
a Avg: average of 10 measurements;
b sd: standard deviation.
Therefore, it was decided to apply the NPOC method for determining the DOC
content (as a measure of the WSOC content) of each aqueous extract of the aerosol
samples. Within the NPOC method, several sparging periods with high purity air were also
tested in order to assess the most appropriate time to remove the IC (as CO2) from the
acidified sample. As shown in Fig. III-3, 2 minutes are proved to be sufficient to remove
Experimental procedures
41
any IC present in the sample. Furthermore, the average peak area count remains almost
constant for times of sparging longer than 2 minutes.
Fig. III-3. Effect of the sparging time in the average peak area counts (grey diamonds, 5 replicas) obtained
in NPOC method using ultrapure water as a sample. The errors bars correspond to the standard
deviation.
Another important aspect when using the NPOC method is the type of container
(glass or plastic) employed to store the HCl reagent, which will be used to acidify the
samples. As shown in Table III-4, when using the HCl stored in plastic containers, the
average peak area count is higher than when using the HCl stored in glass containers. As
such, only HCl stored in glass containers will be used in the subsequent analyses.
Table III-4. Effect of the type of HCl container on the average peak area counts (5 replicas, standard
deviation in brackets) obtained in NPOC method using ultrapure water as a sample.
Type of container
Plastic Glass
Before acidification 257 (63) 254 (24)
After acidification 397 (67) 186 (9)
In this work, the determination of DOC by the NPOC method was performed using
a six point calibration graph, established every week and obtained from standards of
HO2CC6H4CO2K prepared in ultrapure water ranging between 0 mg C L-1
and 1 mg C L-1
.
Fig. III-4 represents an example of a calibration graph used for determining the DOC
content of the aerosol aqueous extracts. Because the confidence interval of intercept does
200
300
400
500
600
0 1 2 3 4
Ave
rage
Pea
k A
rea
Cou
nt
Sparge time (min)
Chapter III
42
not include the zero, each calibration equation was calculated by the algebraic form y = bx
+ a. The standard deviation of the slope Sb and the intercepted Sa ranged between, 6.747-
13.02 and 4.085-7.883, respectively, and the standard error ranged between 12.63 and
24.36. The correlation coefficients range between 0.9986 and 0.9996. The detection limit
of the method (0.016 mg C L-1
) was calculated according to Miller and Miller (2005),
based on the blank (ultrapure water) signal plus three standard deviations of the blank.
Fig. III-4. Example of a calibration graph used for determining the DOC content of the aerosol aqueous
extracts using the NPOC method.
Briefly, the procedure adopted to determine the DOC content started by acidifying
10 mL of each standard/sample with 6 µL of 6M HCl (pH 3) immediately before
sparging with high purity gas for 2 minutes. Then, the standard/sample was injected into
the TC furnace of the analyzer, where it was catalytically oxidized to CO2 according to the
principle of the NPOC method. For each standard and sample, five measurements of DOC
were performed using an injection volume of 25 µL.
3.7. Isolation and fractionation of WSOM from the PM2.5 samples
Prior to the isolation and fractionation of aerosol WSOC, and to obtain sufficient
mass for the subsequent elemental analysis and structural characterization of the WSOM
y = 1306.9x + 182.9
R² = 0.9979
0
400
800
1200
1600
0.0 0.2 0.4 0.6 0.8 1.0
Pea
k A
rea
Co
un
t
TC standard (mg C L-1)
Experimental procedures
43
samples, the aqueous extracts were batched together according to similar features in terms
of ambient conditions and DOC content. These groups are described in section 4.5.
The aerosols WSOC samples were isolated and fractionated using a procedure
similar to that developed by Duarte and Duarte (2005), which is illustrated in Fig. III-5. In
the procedure developed by those authors, an AmberliteTM
XAD-8 resin was used to isolate
the most hydrophobic WSOC components of the aerosol aqueous extracts. However, the
XAD-8 resin is no longer commercially available and it was replaced by the comparable
SupeliteTM
DAX-8 resin, which has been employed to concentrate and isolate DOM from
diverse environments, including the WSOM in this study. The aqueous extracts were
acidified at pH 2.2 with 6M HCl and then pumped onto the DAX-8 resin at a flow rate of
0.8 mL min-1
. After this concentration stage, the organic matter adsorbed onto the resin
was washed with one column volume (35 mL) of ultrapure water, also at a flow rate of 0.8
mL min-1
, to remove the inorganic species. The desalting effluent from the DAX-8 resin
was also collected separately. At a first stage, specifically for aerosols WSOC samples
collected in sampling campaign I, the organic matter retained in the resin was back eluted
with a solution of MeOH:H2O in the proportion 1:1, following the procedure of Duarte and
Duarte (2005). The eluates were evaporated almost to dryness (final volume < 1 mL) in a
rotary evaporator at 30ºC and then re-dissolved in ultra-pure water. Before and after each
step, the UV-Vis absorbance at 250 nm and the DOC content (by the NPOC method) of the
influents and effluents from the DAX-8 column and also of the desalting effluents and
reconstituted eluates were measured for assessing losses of organic matter during the
isolation procedure. Finally, the eluates were freeze-dried and kept on a dessicator over
silica gel for further analysis.
At the end of this first stage, it was verified that the percentage of DOC recovery
from the DAX-8 resin was lower than those reported by Duarte and Duarte (2005) for the
AmberliteTM
XAD-8 resin (additional information in section 4.7.1). In order to improve the
DOC recovery from the DAX-8 resin, two different solutions with different proportions of
MeOH:H2O were tested using a fulvic acid reference sample (Pony Lake, 1R109F)
obtained from the International Humic Substances Society (IHSS). In this study, aqueous
solutions (0.9 -1.25 mg C L-1
) of the Pony Lake fulvic acids were pumped onto the DAX-8
resin. A desalting procedure similar to that previously described was also applied, followed
Chapter III
44
by the back elution of the retained organic matter with solutions of MeOH:H2O in the
following proportions: 1:1 and 3:2. In a similar fashion, the eluates were evaporated almost
to dryness in a rotary evaporator at 30 ºC and then re-dissolved in ultra-pure water. The
results of this study are described in the section 4.7.1, and they indicate that the percentage
of DOC recovery from the DAX-8 resin is higher when employing a higher content of
organic solvent (i.e. MeOH). Therefore, at a second stage, specifically for aerosols WSOC
samples collected in sampling campaign II, the back elution of the retained organic matter
was performed using a solution of MeOH:H2O in the proportion 3:2.
Fig. III-5. Scheme of the isolation/fractionation procedure of the aerosol WSOM samples.
3.8. Elemental analysis of the WSOM from the PM2.5 samples
Elemental analysis of the isolated WSOM was performed with a Truspec 630-200-
200 analyzer. Three replica of the contents of carbon (C), hydrogen (H), and nitrogen (N)
were performed and corrected for the moisture and ash content of the samples, using the
information obtained by thermogravimetric analysis (Duarte, 2006).Usually, the samples
are heated until 750 ºC under an air stream. The temperature program used includes three
steps of heating at 10 ºC min-1
, with a hold time at the final temperature of each step: 60
Experimental procedures
45
min at 60 ºC, 60 min at 100 ºC, and 30 min at 750 ºC. The moisture content refers to the
weight lost after 1h at 60 ºC and the ash content to the final weight at 750 ºC. Taking into
account that at the end of the isolation procedure, very small quantities of the WSOM
fractions were obtained (additional information in section 4.5), in this study it was decided
to use the thermogravimetric data (median values of the percentages of weight loss (at 60
ºC) and ash content) obtained by Duarte (2006) for the aerosol WSOM collected in
Moitinhos, Portugal in order to correct the content of C, H, and N. The oxygen (O) content
of each WSOM sample was calculated by subtraction from 100% (%O = 100-(%C + %H +
%N)) after the previous correction. The atomic ratios H/C, O/C and N/C were also
calculated.
3.9. Spectroscopic characterization of the WSOM from the PM2.5
samples
3.9.1. Ultraviolet- Visible (UV-Vis) spectroscopy
The UV-Vis spectra (in the range of 210–500 nm) were recorded on a UV–Vis
spectrophotometer Shimadzu (Dusseldorf, Kent, Germany) Model UV 2101PC in 1cm
path length quartz cells. Ultrapure water was used as a blank.
3.9.2. Excitation-Emission matrix (EEM) fluorescence spectroscopy
The Emission-Excitation Matrix (EEM) fluorescence spectra were recorded on a
spectrophotometer JASCO (Tokyo, Japan), Model FP-6500. The EEM fluorescence
spectroscopy involved scanning and recording of 20 individual emission spectra (230-510
nm) at sequential increments of 10 nm of excitation wavelength between 220 and 400 nm.
The spectra were recorded at a scan speed of 100 nm min-1
using excitation and emission
slit bandwidths of 10 nm. The peaks due to water Raman scatter were removed from all
spectra by subtracting the ultrapure water blank spectra. In order to avoid concentration
effects, the spectra were also normalized to DOC content of the sample and are shown as
Chapter III
46
specific fluorescence intensity (g-1
C L) versus excitation and emission wavelengths (nm).
Additional corrections for fluctuation of instrumental factors and scattering effects (inner
filter effects) were not applied to the acquired spectra. Nevertheless, all spectra were
recorded using the same instrument and the same experimental conditions, allowing
comparison between them and a qualitative discussion of the fluorescence features.
The synchronous fluorescence spectra presented in this work contain fluorescence
intensity data withdrawn from the EEM fluorescence profiles. The synchronous spectra
with λ of 60 nm were obtained by fitting the mathematical equation λEmission =λExcitation +
λ to the EEM profiles and are also shown as specific fluorescence intensity (g−1
C L)
versus excitation wavelength (nm).
3.9.3. Fourier transform infrared - attenuated total reflectance (FTIR-ATR)
spectroscopy
The FTIR-ATR spectra of the isolated aerosol WSOM samples (section 5.5) were
recorded for frequencies between 550 and 4000 cm-1
in a PerkinElmer FTIR
spectrophotometer (FTIR System Spectrum BX). The spectral resolution was 4 cm-1
and
64 scans were averaged in each spectrum acquisition.
3.9.4. Solid-state cross polarization with magic angle spinning 13
C nuclear magnetic
resonance (CPMAS 13
C NMR) spectroscopy
All CPMAS 13
C NMR spectra were acquired at 125.77 MHz on a Bruker Avance-
500 NMR spectrometer using a standard 4 mm double-bearing probe head. Transients were
recorded with a contact time of 1.5 ms and a spinning rate of 9 kHz. The recycle delay was
5s and the length of the proton 90º pulse was 3.5 s. Chemical shifts are quoted in ppm
from the external calibrant tetramethylsilane.
The solid-state CPMAS 13
C NMR spectra were split into seven spectral regions on
the ppm scale (0–50, 50-60, 60-95, 95-110, 110-160, 160-190, 190-230), each
Experimental procedures
47
corresponding to different carbon functional groups, which are assigned in section 5.6. The
integrated areas of each spectral region were determined off-line using the ACDLabs
software free package. Percentage peak areas of individual peaks were calculated by
dividing their areas by the total spectral peak area of the sample.
3.10. Comprehensive two-dimensional liquid chromatography of
WSOM from PM2.5 samples
The procedure adopted for the analysis of aerosol WSOM samples by means of LC
x LC was based on the method developed by Duarte et al. (2012). The aerosol WSOM
samples were prepared by diluting each sample (before freeze-drying, section 3.7) in 10%
of the mobile phase (v/v) of the 1st dimension. The DOC concentrations of the analysed
WSOM samples (were in the range of 0.1 and 0.4 mg C mL-1
).
The 1st dimension consisted of a JASCO semi-micro HPLC pump (model PU-2085
Plus), a Rheodyne injection valve (model 7725i) equipped with a 20 L loop, and an
Acclaim Mixed-Mode HILIC-1 column (Dionex; diameter 4.6 mm; length 150 mm;
comprised of 5 m high-purity, porous, spherical silica particles with 120 ˚A diameter
pores bonded with alkyl diol functional groups). The 1st dimension was operated in
isocratic mode using a mobile phase composition consisting of 20 mM CH3COONH4 (pH
adjusted to 6.0 with 1.1 mM CH3COOH) and 10% (v/v) ACN. The flow rate was 0.020 mL
min−1
and the temperature of the analytical column was maintained at 30 ºC in a JASCO
column oven (model CO-2065 Plus).
In the 2nd
dimension, a JASCO quaternary low pressure gradient pump (model PU-
2089 Plus) and a PSS Suprema 30 ˚A analytical column (Polymer Standards Service
GmbH; diameter 8 mm; length 150 mm; particle size 10 m; separation range 100–30,000
Da; stationary phase polyhydroxymethacrylate copolymer) were applied. The 2nd
dimension was also operated in isocratic mode with a mobile phase composition consisting
of 20 mM NH4HCO3 (pH 8.0) and 11% (v/v) ACN. The flow rate was 2.5 ml min−1
and
the temperature of the analytical column was also maintained at 30 ºC in a JASCO column
oven. The outlet of the 2nd
dimension column was also connected to three detectors in
Chapter III
48
series: a diode array detector (JASCO, model MD-2010) operating at 254 nm, a
fluorescence detector (JASCO, model FP-2020 Plus) operating at emission/excitation
wavelengths of 240/410 nm and 320/415 nm, and an evaporative light-scattering detector
(SEDEX, model 80-LT-ELSD) operating at 60 C and 3.5 bar.
The 1st and 2
nd dimensions were interfaced with an eight port high pressure two-
position interfacing valve (VICI® AG International) equipped with two identical 50 L
sampling loops. Modulation time was 150 s. The valve was controlled by the PSS WinGPC
Unity software (Polymer Standards Service GmbH) by receiving a start-up signal from a
PSS Universal Data Center (model UDC 810). This software was also used for the
acquisition and handling of all data set. The SEC column in the 2nd
dimension was
calibrated using sodium polystyrenesulfonate standards (MW at peak maximum, (Mp):
208, 891, 4210, 6430, and 15800 Da, obtained from Sigma Aldrich), and HPLC grade
acetone (58 Da) 5% (v/v), which also served as a permeation volume probe. These
standards were prepared by dissolving approximately 1 mg of each compound in 1mL of
the mobile phase of the 2nd
dimension.
49
IV
Global carbon balance and isolation of water-
soluble organic matter from atmospheric aerosols
Chapter IV
50
Global carbon balance and isolation of water-soluble organic matter from atmospheric aerosols
51
4.1. Introduction
This chapter begins with an assessment of the meteorological parameters recorded
during the sampling campaigns I and II in order to group together the atmospheric aerosol
samples according to similar ambient conditions, which will be representative of different
seasonal periods. Then, the average ambient concentrations of the PM2.5, PM2.5-10, TC, OC,
EC, and WSOC for each collected aerosol sample will be evaluated. Based on the average
ambient concentrations of the carbonaceous aerosol fractions, an aerosol mass balance was
also performed in order to infer on its seasonal variability and further assist on the
evaluation of the contribution of primary and secondary sources to fine particulate WSOM.
This chapter also describes the isolation/fractionation procedure employed to obtain a
representative aerosol WSOM fraction of the different seasonal periods.
4.2. Assessment of the meteorological parameters recorded during the
sampling campaigns
The meteorological data, including air temperature, RH, precipitation, and wind
velocity, were provided by the Portuguese Sea and Atmosphere Institute and by the
Department of Physics of the University of Aveiro. The air mass back trajectories were
also computed and evaluated for each sampling campaign. The air mass back trajectories
were obtained by means of the HYbrid Single Particle Lagrangian Integrated Trajectory
(HYSPLIT4) model (24 h interval, 7 day), using the Reanalyses (Global, until October
2010) and the Global Data Assimilation System (GDAS, from November 2010)
meteorological databases, accessed via National Oceanic and Atmospheric Administration
(NOAA) Air Resources Laboratory READY website (Draxler and Rolph, 2013;
http://www.arl.noaa.gov/HYSPLIT.php, last access on March 2013).
Chapter IV
52
4.2.1. Meteorological parameters recorded during sampling campaign I
Table IV-1 summarizes the meteorological data acquired during sampling
campaign I in Summer 2008 and 2009 and in Spring 2009. The median of the air
temperature values were very similar among the different sampling periods, varying from
ca. 15 to 23ºC. The lowest values for the air temperature were recorded during the Spring
2009 season, during the night periods. Also, the total precipitation accumulated was higher
during this somewhat colder season (36.9 mmH2O) than in the warmer periods (3.9 and 0.4
mmH2O for Summer 2008 and 2009, respectively). The widest ranges of variation of the
RH values were also verified for the Spring season, thus yielding the lowest median values
for this meteorological parameter during this period. The median of the wind speed values
were also very similar among the different sampling periods, varying from 1.7 to 2.9 m s-1
.
The air mass back trajectories during the sampling campaign I (Annex A), indicates that
the sampling location is mostly under the influence of wind derived from both maritime
and continental surroundings. The calculated air mass back trajectories also showed that
only two sampling periods in Summer 2008 had air masses travelling most of the time
above the continent. In the sampling periods of 22– 29 July 2008, 01– 08 June 2009, and
27 July – 03 August 2009, the air masses came from the maritime surroundings and
travelled mainly above the Atlantic Ocean.
Global carbon balance and isolation of water-soluble organic matter from atmospheric aerosols
53
Table IV-1. Meteorological data obtained during sampling campaign I (additional information in section 3.3).
Season Sample
Code Sampling date
Air temperature
(°C) Precipitation
b
(mmH2O)
RH (%) Wind velocity (ms-1
) Air mass back
trajectory
Range a Median Range
a Median Range
a Median
Summer
2008
SU08-1 23 -30 Jun 2008 14.4 – 23.0 17.7 0.0 67 - 98 88 0.5 - 7.7 2.6 Continental
SU08-2 08-15 Jul 2008 13.6 – 26.6 18.1 1.0 45 - 94 80 0.3 – 8.4 2.6 Maritime/continental
SU08-3 15-22 Jul 2008 15.7 - 26.6 23.4 0.0 45 - 98 85 0.3 – 5.2 1.7 Continental
SU08-4 22-29 Jul 2008 14.4 - 20.7 20.7 2.9 62 - 98 84 0.6 – 5.4 2.0 Maritime
Spring
2009
SP09-1 27 April – 4 May 2009 8.2 - 26.1 15.2 9.7 31 - 95 76 0.0 – 7.1 2.9 Maritime/continental
SP09-2 04-11 May 2009 12.7 – 25.4 16.4 21.1 33 - 92 76 0.2 – 5.7 2.0 Maritime/continental
SP09-3 18-25 May 2009 9.8 – 20.6 15.7 6.1 52 - 95 79 0.2 – 7.1 2.4 Maritime/continental
SP09-4 25 May – 01 Jun 2009 12.0 – 31.8 19.2 0.0 20 - 95 72 0.1 – 7.4 2.4 Maritime/continental
Summer
2009
SU09-1 01-08 Jun 2009 15.0 – 21.4 17.9 0.2 63 - 95 79 0.3 – 6.7 1.8 Maritime
SU09-2 06-13 Jul 2009 14.6 – 24.5 18.6 0.0 57 - 93 79 0.2 – 7.4 2.7 Maritime/continental
SU09-3 27 Jul – 03 Aug 2009 13.2 – 22.9 19.0 0.1 56 - 94 80 0.1 – 6.6 2.3 Maritime
SU09-4 03-09 Aug 2009 14.7 – 23.0 19.1 0.1 62 - 94 83 0.5 – 5.7 2.4 Maritime/continental
a Maximum and minimum values recorded during each sampling period.
b Total precipitation accumulated in each sampling period.
Chapter IV
54
4.2.2. Meteorological parameters recorded during sampling campaign II
The weekly and seasonal variability, in terms of median, minimum, and maximum
values, of the air temperature, RH, wind velocity, and total precipitation accumulated is
shown in Fig. IV-1. As expected, the highest median values for the temperature (Fig. IV-
1a) were verified during the Summer season (the highest value, 39ºC, was recorded in
week AVE 33), although higher median values were also verified in weeks AVE 23, 25,
and 27 during Spring 2010 and weeks AVE 40 to 43 in Autumn 2010. By the end of week
AVE 46, in Autumn 2010, the temperature gradually started to decrease, reaching a
minimum of 0 ºC in week AVE 4 in mid December 2009 and of 2 ºC in weeks AVE 50
and 58, in early December 2010 and late January 2011, respectively. In what concerns the
annual distribution of the RH (Fig. IV-1c), the range of variation was between 20% and
99%. Typically, and regardless of the seasonal period, the values of the RH follow a daily
pattern, with higher values during the night, coincident with lower air temperatures, as can
be observed in Fig. IV-2. Regarding the median values of the wind velocity (Fig. IV-1d),
values lower than 3 ms−1
were observed across most of the sampling period, although
during weeks AVE 20 and 22 in Spring and week AVE 61 in Winter 2011 it reached 4
ms−1
.
Global carbon balance and isolation of water-soluble organic matter from atmospheric aerosols
55
Fig. IV-1. Weekly and seasonal variability (in terms of median, minimum, and maximum values) of the (a) air temperature (in ºC), (b) total precipitation accumulated
(in mmH2O), (c) RH (in %), and (d) wind velocity (in ms-1
) recorded during the sampling campaign II.
Chapter IV
56
Fig. IV-1. Continued.
Global carbon balance and isolation of water-soluble organic matter from atmospheric aerosols
57
Fig. IV-2. Daily variability of the air temperature (in ºC) and HR (in %) recorded between 1 and 4 January
2010.
Fig. IV-1b illustrates the weekly variation of the total accumulated precipitation
recorded during the sampling campaign II. As shown, the occurrence of precipitation was
much lower in the Summer season (total of 2 mmH2O between weeks AVE 28 and AVE
39) than in the colder seasons (ranging from 0 to 173 mmH2O), with the highest value
being recorded in week AVE 50 (in Autumn 2010).
The air mass backward trajectories were also calculated for each week of the
sampling campaign II. The entire record of air mass backward trajectories during this
sampling campaign can be found in Annex C. Fig. IV-3 presents the predominant air mass
back trajectories in each season (Autumn 2009, Winter 2010, Spring 2010, Summer 2010,
Winter 2011, and Spring 2011). As shown in Fig. IV-3, during Autumn 2009 and Autumn
2010, the sampling location is mostly under the influence of wind derived from maritime
surroundings, whereas in Winter 2010, Spring 2010, Winter 2011 and Spring 2011 the air
masses come from both maritime and continental surroundings. The calculated air mass
back trajectories also show that in Summer 2010, the air masses were travelling most of the
time above the continent. Additionally, during this season, the occurrence of massive forest
fires episodes was registered, with the largest incidence being verified between weeks 33
and 35, coincident with the period of highest air temperature (maximum of 39 °C).
According to the report of the Agriculture, Rural Development and Fisheries Ministry
(MADRP-Direcção de Unidade de Defesa da Floresta, 2011), the total burnt area of forest
land was about 133 090 ha in 2010 at Portugal Continental, with 388 “Hot flashes” and 41
fires occurring between weeks 28 and 32, and 972 “Hot flashes” and 79 fires occurring
Chapter IV
58
between weeks 33 and 35 at District of Aveiro (total burnt area of forest land was about 8
299 ha). In Winter 2010, Winter 2011 and Spring 2011, the air masses trajectories came
from both maritime and continental surroundings, although sporadically they also travelled
above North Africa.
Fig. IV-3. Air mass backward trajectories ending at Aveiro at distinct altitudes (>500 m a.g.l.) during:
Autumn 2009 (AVE 1, 23–30 November 2009); Winter 2010 (AVE 10, 2-9 February 2010);
Spring 2010 (AVE 20, 27 April to 4 May 2010); Summer 2010 (AVE 35, 09–16 August 2010);
Autumn 2010 (AVE 46, 2-9 November 2011), Winter 2011 (AVE 64, 8-15 March 2011), and
Spring 2011 (AVE 66, 22-29 March 2011).
Global carbon balance and isolation of water-soluble organic matter from atmospheric aerosols
59
Fig. IV-3. Continued.
4.3. Assessment of OC adsorption phenomena onto quartz filters during
aerosol sampling
The most common approach for determining the composition of atmospheric
aerosols involves the collection of the particulate matter onto filter substrates. The
sampling time depends on the analytical sensitivity of the different methodologies, and it
can vary typically from several hours in urban areas to a day or more under clean
background conditions or in areas with very low particle concentrations (Jacobson et al.,
2000; McMurry, 2000). The collection of aerosol particles by means of a high-volume
sampler using quartz fibre filters is known to be subjected to potential sampling bias. From
Chapter IV
60
one side, the volatilization of SVOC has been referred as a significant source of negative
bias, whereas the adsorption of gas-phase compounds onto filters during and/or after
sample collection is considered as a source of positive bias (Turpin et al., 1994). The
occurrence of such phenomena has been reported in several studies (Mayol-Bracero et al.,
2002; Viana et al., 2006a; Salma et al., 2007; Mkoma et al., 2010) and it can affect the
composition of the collected aerosols particles in relation to the real atmosphere
composition when filter-based devices are used. Negative bias during aerosol particles
sampling can be estimated using carbon-impregnated in backup glass filter approach
(Subramanian et al., 2004), whereas positive bias can be assessed using a tandem filter
method (Turpin et al., 1994). This latter method consists of (1) a Teflon filter followed by
a quartz filter or (2) a set of two (front and back) quartz fibre filters (Turpin et al., 1994).
According to comprehensive review conducted by Turpin et al. (2000), sampling bias
contributions to OC mass measurements range from –80% for volatilization-induced bias
up to +50% for adsorption-induced bias.
Several papers have also reported that the occurrence of volatilization and/or
adsorption phenomena during aerosol sampling are of importance for the WSOC
component (Mayol-Bracero et al., 2002; Salma et al., 2007). Using quartz fibre filters
tandem method for aerosol sampling in Amazonia (forest environment), Mayol-Bracero et
al. (2002) reported back-to-front filters ratios for the WSOC concentration of about 5%.
Salma et al. (2007) have also studied sampling bias on an urban environment using the
tandem quartz fibre filter method. The authors reported an average back-to-front ratio of
28% for the WSOC, which was even higher than that estimated for the OC component
(17%). These results allowed the authors to suggest that the organic compounds that are
mainly responsible for the positive (adsorptive) sampling bias are likely to be more water
soluble, thus they possibly contain more polar functional groups or they may be more
oxidized than the organic matter in general (Salma et al., 2007).
Taking into account the results reported in the literature for the occurrence of
sampling artefacts when using filters-based devices, this work also entailed an assessment
of the positive sampling artefacts (using the tandem quartz fibre filter method) eventually
existing during aerosol collection at Aveiro (sampling campaign I). The PM2.5 and PM2.5-10
mass concentrations (g m-3
) from both the front and back filters are shown in Fig. IV-4.
Global carbon balance and isolation of water-soluble organic matter from atmospheric aerosols
61
Fig. IV-4. Distribution of PM2.5 and PM2.5-10 mass concentrations (g m-3
) on the front and back filters
during sampling campaign I.
The weekly PM2.5 concentrations in the top filter varied from 8.2 to 13.5, 7.0 to
17.1 and 8.9 to 10.5 g m-3
for Summer 2008, Spring 2009 and Summer 2009 seasons,
respectively. Regarding the PM2.5-10 fraction, the mass concentrations in the front filters
ranged from 8.2 to 13.5, 8.6 to 13.3 and 7.8 to 11.1 g m-3
for Summer 2008, Spring 2009
and Summer 2009, respectively. The back-to-front ratios for the PM2.5 concentration
ranged from 5.3 to 59.1, 2.5 to 37.4, and 5.6 to 21.2 % for the Summer 2008, Spring 2009
and Summer 2009 seasons, respectively, whereas the back-to-front ratio for the PM2.5-10
fraction and for the same seasons ranged from 4.0 to 22.7, 5.0 to 13.6 and 5.2 to 14.1 %,
respectively.
The average concentrations (µg C m-3
) and standard deviations for the OC, EC, and
WSOC fractions in the front and back filters from both PM2.5 and PM2.5-10 samples are
Samples
SU08-1
SU08-2
SU08-3
SU08-4
SP09-1
SP09-2
SP09-3
SP09-4
SU09-1
SU09-2
SU09-3
SU09-4
PM2.5
conc
entra
tion (
gm-3
)
0
2
4
6
8
10
12
14
16
18
Front filters
Back filters
Summer 2008 Spring 2009 Summer 2009
Samples
SU08-1
SU08-2
SU08-3
SU08-4
SP09-1
SP09-2
SP09-3
SP09-4
SU09-1
SU09-2
SU09-3
SU09-4
PM2.5
-10 c
once
ntrati
on (
gm-3
)
0
2
4
6
8
10
12
14
Front filters
Back filters
Summer 2008 Spring 2009 Summer 2009
Chapter IV
62
summarized in Table IV-2. The OC concentrations in the front filters varied from an
average of 0.94 (0.09) to 1.98 (0.17) μg C m−3
for the PM2.5 samples and 0.28 (0.03) to
0.51 (0.04) μg C m−3
for the PM2.5-10 samples, with the samples collected in the Spring
season presenting higher average values than those collected during the warmer periods.
Table IV-2. Average concentration (in µg C m-3
) and standard deviation (in brackets) of the main
carbonaceous components from both PM2.5 and PM2.5-10 in the top and back filters during
sampling campaign I. The number of aerosol samples in each season was 4.
Season / PM
OC (µg C m-3
) EC (µg C m-3
) WSOC (µg C m-3
)
Front Back Front Back Front Back
PM2.5
Summer 2008 1.21 (0.13) 0.41 (0.06) 0.31 (0.14) n.d. 0.64 (0.10) 0.17 (0.03)
Spring 2009 1.98 (0.17) 0.35 (0.03) 0.08 (0.02) n.d. 0.85 (0.06) 0.28 (0.04)
Summer 2009 0.94 (0.09) 0.19 (0.03) 0.13 (0.01) n.d. 0.43 (0.03) 0.14 (0.01)
PM2.5-10
Summer 2008 0.48 (0.06) 0.02 (0.02) 0.02 (0.03) n.d. 0.12 (0.01) n.a.
Spring 2009 0.51 (0.04) 0.03 (0.01) 0.02 (0.02) n.d. 0.12 (0.04) n.a.
Summer 2009 0.28 (0.03) 0.02 (0.00) 0.01 (0.00) n.d. 0.08 (0.01) n.a.
n.d.: not detected; n.a.: not assessed
These results suggest that during Summer, the higher temperatures favor the
gaseous phase in the gas-particle partitioning of the SVOC. Moreover, the highest OC
concentrations were found in the fine size range particles, as should be expected from the
typical conceptual segmentation models of chemical species versus the aerosol mass size
distribution (Seinfeld, 1986; Krivácsy and Molnár, 1998; Seinfeld and Pandis, 1998;
Samara and Voutsa, 2005).
Overall, the back-to-front ratios for the OC in the PM2.5 samples ranged from 15 to
47%, 11 to 38%, and 17 to 21% in Summer 2008, Spring 2009, and Summer 2009 seasons,
respectively. For the same seasons, but in the coarse particles, the back-to-front ratios for
the OC are much lower, ranging from 2.6 to 8.4%, 4.4 to 9.4%, and 6.7 to 9.6%,
respectively. The EC concentration in front filters varied from an average of 0.08 (0.02) to
Global carbon balance and isolation of water-soluble organic matter from atmospheric aerosols
63
0.31 (0.14) μg C m−3
for the PM2.5 samples and 0.01 (0.00) to 0.02 (0.03) μg C m−3
for the
PM2.5-10 samples. Since EC has predominantly primary origin (Seinfeld and Pandis, 2006),
no EC was detected in the back filters. This finding is in accordance with those of reported
by Viana et al (2006a), Salma et al. (2007), Mkoma et al. (2010). With regard to the
WSOC in the front filters of the PM2.5 samples, this water-soluble fraction accounts, on
average, to (539)%, (423)%, and (4611)% of the OC for the Summer 2008, Spring
2009, and Summer 2009 seasons, respectively. In the PM2.5-10 samples, also in the front
filters, the contribution of WSOC to the OC is somewhat lower, accounting on average to
(249)%, (234)%, and (287)% for the Summer 2008, Spring 2009, and Summer 2009
seasons, respectively. When looking at the back-to-front ratios for the WSOC in the PM2.5
samples, the obtained values ranged from 16 to 39%, 19 to 68%, and 24 to 53% in Summer
2008, Spring 2009, and Summer 2009, respectively. With the exception for the Summer
2008 season, the back-to-front ratios for the WSOC in the PM2.5 samples are higher than
those for OC. These results suggest that the organic compounds that are responsible for the
positive sampling bias (adsorptive phenomenon) seem to be more water-soluble in nature.
These findings are in agreement with those of Salma et al. (2007), who reported back-to-
front ratios of 28% for the WSOC fraction and of 17% for the OC in PM2.5 samples
collected in an urban area, during a non-heating Spring season. These results also suggest
that during sampling the water-soluble organic compounds in the back filters were formed
via adsorption on the filter surface of volatile gaseous precursors from both anthropogenic
and biogenic emission sources, and by heterogeneous reactions with oxidants (Salma et al.,
2007).
Several methods have been proposed to correct the sampling bias phenomena when
collecting atmospheric aerosols, including the backup filter subtraction to estimate the
positive bias (e.g. Viana et al., (2006a) and Salma et al., (2007)), and the use of denuder-
based methods to correct positive and negative (volatilization) bias (Turpin et al., 2000).
Recently, Maimone et al. (2011) have discussed bias correction options in large routine
sampling networks, and the authors presented a regression method based on the
assumption of a linear relationship between the measured OC and the mass of particulate
matter, whose linearity are not always valid. The authors concluded that, taking into
account that the adsorption of organic gases increases with increasing atmospheric
concentrations of organics products, the subtraction of an average bias from all samples
Chapter IV
64
will underestimate OC for lower-concentration samples and will overestimate OC for
higher-concentration samples (Maimone et al., 2011).
Apparently, one of the easiest ways for obtaining sampling-positive bias-corrected
atmospheric concentrations of OC and WSOC is by subtraction of their respective amounts
determined on the back quartz filters. Nevertheless, for applying this method one must
assume that the back-to-front ratios for both OC and WSOC are independent of the
seasonal period in order to apply the same correction factor for the all set of atmospheric
aerosol samples. This assumption seems difficult to accomplish since both OC and WSOC
concentrations in the back quartz fibre filters vary in accordance to the meteorological
parameters (i.e., season). Therefore, in this work, and for the sampling campaign II, the
collection of atmospheric aerosol samples was performed without any attempt for
controlling adsorption/desorption phenomena on the filter, because of practical difficulties
in applying methodologies such as backup filter subtraction, denuder for pre-removal of
gaseous oxidants and SVOCs, and foam plugs for post-filter collection of volatilized
particulate components from filter surfaces, with extended high-volume sampling.
Consequently, volatilisation/condensation processes of SVOCs are likely to occur on the
filter or on particles surface. Also, oxidation of filter deposited organics by strong oxidants,
such as ozone, may happen during filtration. Therefore, the measured concentrations for
oxygenated organic species, WSOC in particular, represent an upper limit of the true
atmospheric levels (Pio et al., 2001).
4.4. Seasonal trend of fine and coarse aerosol in Aveiro
Fig. IV-5 shows the weekly ambient concentrations of both PM2.5 and PM2.5-10
samples collected in sampling campaign II. Due to technical reasons and collection of field
blank (as already referred in section 3.3), only four samples were collected during the
Winter season in 2010. It is possible to verify that the highest concentrations of particulate
matter predominate in the fine size fraction of the aerosol.
Global carbon balance and isolation of water-soluble organic matter from atmospheric aerosols
65
Fig. IV-5. PM2.5 and PM2.5-10 concentrations (in μg m-3
) during the annual sampling campaign II in Aveiro.
Weeks/Samples
AV
E 1
AV
E 2
AV
E 3
AV
E 4
AV
E 5
AV
E 6
AV
E 7
AV
E 8
AV
E 9
AV
E 1
0A
VE
11
AV
E 1
2A
VE
13
AV
E 1
4A
VE
15
AV
E 1
6A
VE
17
AV
E 1
8A
VE
19
AV
E 2
0A
VE
21
AV
E 2
2A
VE
23
AV
E 2
4A
VE
25
AV
E 2
6A
VE
27
AV
E 2
8A
VE
29
AV
E 3
0A
VE
31
AV
E 3
2A
VE
33
AV
E 3
4A
VE
35
AV
E 3
6A
VE
37
AV
E 3
8 A
VE
39
AV
E 4
0A
VE
41
AV
E 4
2A
VE
43
AV
E 4
4A
VE
45
AV
E 4
6A
VE
47
AV
E 4
8A
VE
49
AV
E 5
0A
VE
51
AV
E 5
2A
VE
53
AV
E 5
4 A
VE
55
AV
E 5
6 A
VE
57
AV
E 5
8 A
VE
59
AV
E 6
0A
VE
61
AV
E 6
2A
VE
63
AV
E 6
4A
VE
65
AV
E 6
6
Par
ticu
late
Mat
ter
Conce
ntr
atio
n (
g m
-3)
0.0
10.0
20.0
30.0
40.0
PM2.5
PM2.5-10
Autu
mn 2
00
9
Winter 2010 Spring 2010 Summer 2010 Autumn 2010 Winter 2011
Spri
ng 2
01
1
Chapter IV
66
Overall, the total aerosol mass ranged from 8 to 37 μg m−3
and from 4 to 30 μg m−3
for PM2.5 and PM2.5-10 fractions, respectively, with the highest concentrations being
observed during the colder periods (Winter 2010, Autumn 2010, and Winter 2011). The
annual trend of total aerosol mass also shows an increase in the levels of both PM2.5 and
PM2.5-10 fractions in Summer 2010, particularly when compared with the levels obtained
during Spring 2010. As it will be explained in the following paragraphs, this increase in the
total aerosol mass could be associated with forest fire events occurring in Summer 2010,
which are an important source of atmospheric particles in the region.
Similar annual trends for the PM2.5 mass concentrations (except for Summer
season) have also been reported in the literature for other urban locations: based on daily
sampling in the urban area of Xiamen, China, the average PM2.5 concentrations varied in
the order of Summer < Spring < Autumn < Winter (Zhang et al., 2011); in three European
cities (Amsterdam, Barcelona, and Ghent), the PM2.5 mass was higher in the Winter and
lower in the Summer seasons (Viana et al., 2007). The results for PM2.5 mass
concentrations obtained in this study also show a similar trend to those observed by Duarte
et al., (2007) in a rural area, with high agricultural activity, located near the city of Aveiro.
Although a rural site typically refers to a site distanced from population centers, roads and
industrial areas, thus having low particulate matter concentrations, the results of this study
suggests that there is no clear difference on the levels of PM2.5 between urban and rural
areas in the region of Aveiro.
The PM2.5-10 mass concentrations do not seem to follow a seasonal trend similar to
that of PM2.5. Indeed, the PM2.5-10 mass concentrations are lower in the colder periods
(namely, in Spring 2010, Autumn 2010, and Winter 2011) than in the warmer period
(Summer 2010). The coarse particles have a short residence time in the atmosphere and are
rapidly removed from the air by sedimentation. This feature alongside the high levels of
precipitation accumulated during the colder periods, as shown in Fig. IV 1b, may have
promoted the removal of the coarse particles from the atmosphere, thus reducing its
concentration in the atmosphere. The higher levels of PM2.5-10 found in the warmer period
could be related to local emission sources, such as industrial activities, intense traffic load,
and wind driven re-suspension of road dust, that could considerably contribute to enhance
the levels of coarse particles in the atmosphere. During the Summer season, particularly
Global carbon balance and isolation of water-soluble organic matter from atmospheric aerosols
67
between weeks AVE 33 and AVE 35, the unusual high total aerosol masses (PM2.5-10 +
PM2.5) could also be explained by the emissions from the forest fires that took place in
those periods (referred in section 4.2.2). For the weeks AVE 33, AVE 34, and AVE 35, the
air mass backward trajectories (in Annex A) indicate the continental transport of the smoke
from different inland regions of the Iberian Peninsula to the western coastal sampling site
of Aveiro, thus promoting an increase of the total aerosol mass collected in these periods.
These findings are corroborated by those of Pio et al (2008), also for the region of Aveiro,
during Summer 2003, when unusually large forested areas (>300,000 ha) were destroyed
by fire. The authors reported that during these wildfire episodes, the particulate matter
levels were elevated 2- to 3-fold above the prior and post smoke baseline periods.
During the Spring and Summer seasons in sampling campaign II, Portugal, as well
the rest of Europe, was under the influence of the transport and input of dust derived from
natural sources, including the dust from the Sahara desert (5 to 19 July 2010) and the ashes
from the Eyjafjallajökull volcano eruption in Iceland (between April and May 2010). As an
example, Fig. IV-6 shows a map of the dispersion of the volcano ashes in the European
airspace at May 9th
2010, which correspond to week AVE 21 (Spring 2010).
Fig. IV-6. Map of plume of ash trajectories emitted from the Iceland Eyjafjallajökull volcano eruption in 9th
May 2010, obtained from VAAC at the Meteorological Office London website
(http://www.metoffice.gov.uk/aviation/vaac/vaacuk_vag.html).
Chapter IV
68
The map was obtained from the Volcanic Ash Advisory Centre (VAAC) at the
Meteorological Office London website. The Eyjafjallajökull volcano firstly erupted on
April 14th
2010 and a second eruption was verified in May 2010, emitting great amounts of
particulate matter into the atmosphere. Due to atmospheric conditions over Iceland, the
plume of ashes was transported towards western European, leading to the close of most of
the European airspace. It is expected that the emissions from these natural sources could
contribute to an increase in the levels of particulate matter in the atmosphere. For example,
according to Remoundaki et al. (2012), an intense Sahara dust storm was responsible for
the highest PM2.5 concentration (100 μg m-3
) recorded in the urban environment of Athens,
where the Sahara dust accounted for 96% of the collected PM2.5 mass. Navrátil et al.
(2012) have demonstrated that the deposition of particulate matter in Prague-Suchdol due
to dust deposition in the period affected by volcanic ash, increased gradually from 0.001 g
m-2
day-1
on 14 (14–15) April 2010 to 0.150 g m-2
day-1
on 19 (19–20) April, with an
excessive maximum of 0.600 g m-2
day-1
on 21 (21–22) April. It should be mentioned,
however, that the assessment of the impact of the aforementioned natural events (volcano
eruption and Sahara dust) on the total amount of atmospheric particles in Aveiro was not
verified in this study and, therefore, no conclusions can be withdrawn regarding this issue.
Fig. IV-7 shows the weekly variation of the PM2.5-to-PM10 ratio (PM2.5/PM10,
where PM10 is calculated as PM2.5 + PM2.5-10) for the samples collected during sampling
campaign II. As shown, the majority of the PM2.5/PM10 ratios are higher than 50%. A few
exceptions are seen for week AVE 7 in Winter 2010, as well as for week AVE 27 in Spring
2010, weeks AVE 30 to AVE 32 and AVE 34 in Summer 2010, and weeks AVE 40 to
AVE 42 in Autumn 2010, where the PM2.5/PM10 ratios are between 30 and 50%, thus
implying that the PM2.5-10 fraction prevailed over the PM2.5 fraction at this sampling site.
Global carbon balance and isolation of water-soluble organic matter from atmospheric aerosols
69
Fig. IV-7. Weekly variation of the PM2.5/PM10 ratio during sampling campaign II.
According to Harrison et al. (2012), high PM2.5/PM10 ratios (0.4-0.9) are likely due
to a higher contribution of particulate matter with a markedly secondary origin, in urban
background sites across UK. PM2.5/PM10 ratios values higher than 50% were also found by
Lin et al. (2009) for an urban sites in central Taiwan (Taichun City: average of 0.56 ± 0.09
and Changhwa: average of 0.57 ± 0.11, respectively) during the Winter and Spring
sampling periods. At a coastal site, Wuchi, the authors observed a PM2.5/PM10 ratio value
of 0.48 ± 0.12. The authors attributed this ratio to both high contribution of sea salt aerosol
and suspended dust under the strong wind speed conditions during sampling. PM2.5/PM10
ratios ranging from 0.44 to 0.90 were found across Europe and no general relationship can
be established to the location (site type –urban or rural) or to its geographical location in
the continent (Putaud et al., 2010). In the present study, PM2.5/PM10 ratios lower than 50%
in Summer 2010 are probably due to the occurrence of forest fire events. Another
explanation may be related to the proximity of the sampling site to the sea and,
consequently, the greater influence of sea spray particles, which are likely to predominate
in the coarse fraction of the total particulate matter in the atmosphere.
Fig. IV-8 shows the weekly variation of the total PM2.5-10 mass concentrations as
well as the average concentrations (and the standard deviation) of TC, EC, OC, and WSOC
in the PM2.5-10 samples collected during sampling campaign II. The concentrations of TC
Weeks/Samples
AV
E 1
AV
E 2
AV
E 3
AV
E 4
AV
E 5
AV
E 6
AV
E 7
AV
E 8
AV
E 9
AV
E 1
0A
VE
11
AV
E 1
2A
VE
13
AV
E 1
4A
VE
15
AV
E 1
6A
VE
17
AV
E 1
8A
VE
19
AV
E 2
0A
VE
21
AV
E 2
2A
VE
23
AV
E 2
4A
VE
25
AV
E 2
6A
VE
27
AV
E 2
8A
VE
29
AV
E 3
0A
VE
31
AV
E 3
2A
VE
33
AV
E 3
4A
VE
35
AV
E 3
6A
VE
37
AV
E 3
8
AV
E 3
9A
VE
40
AV
E 4
1A
VE
42
AV
E 4
3A
VE
44
AV
E 4
5A
VE
46
AV
E 4
7A
VE
48
AV
E 4
9A
VE
50
AV
E 5
1
AV
E 5
2A
VE
53
AV
E 5
4
AV
E 5
5
AV
E 5
6
AV
E 5
7A
VE
58
AV
E 5
9A
VE
60
AV
E 6
1A
VE
62
AV
E 6
3A
VE
64
AV
E 6
5A
VE
66
PM
2.5
/PM
10 (
%)
30
40
50
60
70
80
90
Autu
mn 2
00
9
Winter 2010 Spring 2010 Summer 2010 Autumn 2010 Winter 2011
Sp
ring 2
01
1
Chapter IV
70
and OC follow a similar seasonal trend, with maximum values during Autumn and Winter
seasons and minimum concentrations during the warmer periods (Summer and Spring
seasons). The OC is the predominant component of the carbonaceous material in the PM2.5-
10 samples, accounting for more than 69% of the TC, while the EC and WSOC contribution
to the TC content of the PM2.5-10 samples was in the range of 0 to 33% and 4 to 20%,
respectively. The EC concentration for PM2.5-10 fraction varied from 0 to 0.175 μg C m-3
during the sampling campaign II (Fig. IV-8). The concentrations of WSOC in the PM2.5-10
samples were an average of in the range of 0.024 (0.001) – 0.161 (0.008) μg C m-3
, with
the higher concentrations being found in the Autumn periods (0.042 (0.003) – 0.135
(0.005) μg C m-3
), except the highest one which was found in Summer (AVE 38- 0.161
(0.008) μg C m-3
), and the lowest values in the Spring 2010 season (0.024 (0.001) – 0.087
(0.006) μg C m-3
).
The average concentrations of WSOC in the coarse particles collected in Summer
2010 (0.069 (0.018) μg C m-3
) and Winter 2011 (0.047 (0.012) μg C m-3
) seasons were
found to be lower than those reported for a coastal urban site in China (1.1 μg C m-3
in
Summer and 1.9 μg C m-3
in Winter) (Huang et al., 2006). Also, the average concentrations
of WSOC obtained in the present study during the Summer season are lower than those
found in the city of Oporto (0.20 (0.70) μg C m−3
) for particles within the 3.0–10 μm size
range collected in the period between August and September 2004 (Duarte et al., 2008a).
These differences are likely be related to the contribution of the different dominant sources
to the concentration levels of the WSOC aerosol component.
The contribution of WSOC to OC (WSOC/OC) in the PM2.5-10 fraction varied
between 4 and 20%, with higher average values being found during Summer 2010 (13
(4)%). For weeks AVE 33 to AVE 35, corresponding to the high forest fire pollution event,
the WSOC concentration ranged between 0.06 and 0.08 μg C m-3
, which could be
suggested to a contribution from wood combustion and the average WSOC/OC ratio
during these week was low (11%). The average of WSOC/OC ratio of the aerosol samples
observed during Summer 2010 (13 (4) %) was in the range of the data reported for coastal-
rural site (13.2 (±4.6)%) and lower than those found for urban area (17.6 (±2.7)%) during
Summer 2004 for particles within the coarse size range (3.0–10 μm) (Duarte et al., 2008a).
Global carbon balance and isolation of water-soluble organic matter from atmospheric aerosols
71
Fig. IV-8. Weekly variation of the PM2.5-10 mass concentration (in µg m
-3) and average concentrations of the carbonaceous fractions (in µg C m
-3) of the PM2.5-10
samples collected during sampling campaign II. Error bars refers to standard deviation.
Weeks/Samples
AV
E 1
AV
E 2
AV
E 3
AV
E 4
AV
E 5
AV
E 6
AV
E 7
AV
E 8
AV
E 9
AV
E 1
0A
VE
11
AV
E 1
2A
VE
13
AV
E 1
4A
VE
15
AV
E 1
6A
VE
17
AV
E 1
8A
VE
19
AV
E 2
0A
VE
21
AV
E 2
2A
VE
23
AV
E 2
4A
VE
25
AV
E 2
6A
VE
27
AV
E 2
8A
VE
29
AV
E 3
0A
VE
31
AV
E 3
2A
VE
33
AV
E 3
4A
VE
35
AV
E 3
6A
VE
37
AV
E 3
8A
VE
39
AV
E 4
0A
VE
41
AV
E 4
2A
VE
43
AV
E 4
4A
VE
45
AV
E 4
6 A
VE
47
AV
E 4
8 A
VE
49
AV
E 5
0A
VE
51
AV
E 5
2A
VE
53
AV
E 5
4A
VE
55
AV
E 5
6A
VE
57
AV
E 5
8A
VE
59
AV
E 6
0A
VE
61
AV
E 6
2A
VE
63
AV
E 6
4A
VE
65
AV
E 6
6
PM
2.5
-10 C
once
ntr
atio
n (
g m
-3)
0
4
8
12
16
20
24
28
32
Car
bon
Co
nce
ntr
atio
n (
g C
m-3
)
0.0
0.5
1.0
1.5
2.0
2.5
PM2.5-10
TC
OC
EC
WSOC
Chapter IV
72
In general, the seasonal distribution of the average ambient concentrations of PM2.5,
TC, OC, EC, WSOC, and WINSOC (water-insoluble organic carbon,
(WINSOC=OC−WSOC)) are shown in Fig. IV-9. The concentrations of TC, OC, and
WSOC showed a similar trend, with minimum values in Spring and in the first weeks of
Summer (AVE 28 to AVE 31) and maximum concentrations in Autumn and Winter.
During sampling campaign II, the OC was the predominant fraction of the carbonaceous
material in the PM2.5 samples, accounting for more than 57% of the TC, while EC was
found to have only a minor contribution to the TC content during Autumn 2009 (0.29-
0.49%), and Winter 2011 (1.28-3.85%). For Spring and Summer periods, the contribution
of EC to TC ranged from 16.06 to 42.48% and 11.60 to 42.73%, respectively. OC
concentrations in PM2.5 ranged from 0.57 to 13 μg C m−3
, and they were systematically
higher than those found in PM2.5-10. Regarding the EC levels in PM2.5, the concentrations
ranged from 0.02 to 1.44 μg C m−3
, being also higher than those found in PM2.5-10. During
the whole sampling period of campaign II, the EC concentration in PM2.5 reached higher
values in Autumn 2010 (ranging from 0.506 (0.021) to 1.441 (0.031) μg C m-3
), and lower
values in Autumn 2009 (ranging from 0.0245 (0.0068) to 0.028 (0.010) μg C m-3
).
The average EC concentration measured in this work for the summer season (0.69
(0.07) μg C m-3
) was found to be higher than that reported by Duarte et al. (2007) for a
coastal- rural area near the city of Aveiro (0.45 (0.02) μg C m-3
). However, for the Autumn
and Winter seasons, the authors reported higher average EC concentrations (1.27 (0.22) μg
C m-3
and 1.30 (0.44) μg C m-3
, respectively) than those obtained in this study for Autumn
2010 (0.881 (0.113) μg C m-3
) and for Winter 2011 (0.144 (0.045) μg C m-3
). It was also
verified that the levels of EC increased for samples collected in weeks AVE 33 to AVE 35
((0.977 (0.042) – 1.10 (0.027) μg C m-3
) (during the forest fire events in the Summer
season, although the highest values were obtained for samples collected during Autumn
2010 season (0.507 (0.021) - 1.441 (0.031) μg C m-3
). Surprisingly, the lowest levels of EC
were found for samples collected during the Autumn 2009 (0.0245 (0.0068) – 0.028
(0.010) μg C m-3
).
Global carbon balance and isolation of water-soluble organic matter from atmospheric aerosols
73
Fig. IV-9. Weekly variation of the PM2.5 mass concentration (in µg m
-3) and average concentrations of the carbonaceous fractions (in µg C m
-3) of the PM2.5 samples
collected during sampling campaign II. Error bars refers to standard deviation.
Weeks/Samples
AV
E 1
AV
E 2
AV
E 3
AV
E 4
AV
E 5
AV
E 6
AV
E 7
AV
E 8
AV
E 9
AV
E 1
0A
VE
11
AV
E 1
2A
VE
13
AV
E 1
4A
VE
15
AV
E 1
6A
VE
17
AV
E 1
8A
VE
19
AV
E 2
0A
VE
21
AV
E 2
2A
VE
23
AV
E 2
4A
VE
25
AV
E 2
6A
VE
27
AV
E 2
8A
VE
29
AV
E 3
0A
VE
31
AV
E 3
2A
VE
33
AV
E 3
4A
VE
35
AV
E 3
6A
VE
37
AV
E 3
8A
VE
39
AV
E 4
0A
VE
41
AV
E 4
2A
VE
43
AV
E 4
4A
VE
45
AV
E 4
6 A
VE
47
AV
E 4
8 A
VE
49
AV
E 5
0A
VE
51
AV
E 5
2A
VE
53
AV
E 5
4A
VE
55
AV
E 5
6A
VE
57
AV
E 5
8A
VE
59
AV
E 6
0A
VE
61
AV
E 6
2A
VE
63
AV
E 6
4A
VE
65
AV
E 6
6
PM
2.5 C
onc
entr
atio
n (
g m
-3)
0
10
20
30
40
Car
bon
Con
cent
rati
on (
g C
m-3
)
0
2
4
6
8
10
12
14
16
18
PM2.5
TC
OC
EC
WSOC
WINSOC
Aut
umn
20
09
Winter 2010 Spring 2010 Summer 2010 Autumn 2010 Winter 2011
Sp
ring
20
11
Chapter IV
74
A number of other studies had measured the EC levels of PM2.5 at urban area. For
example: Viidanoja et al (2002) observed an average EC concentration of 0.6 (0.2), 1.3
(0.7) and 2.1 (1.3) for warmer months (July 2000, August 2000 and September 2000,
respectively) and 1.1 (0.4), 1.1 (0.3), and 1.0 (0.4) for colder months (December 2000,
January 2001, and February 2001, respectively) in an urban area in Helsinki, Finland; Kim
et al (2011) found an average EC concentration of 2.32 (0.44), 3.69 (0.38), 3.88 (0.79), and
2.14 (0.69) (µg m-3
) for summer 2009, fall 2009, Winter 2010 and Spring 2010,
respectively in an urban area of Japan (Saitama City); Zhang et al (2011) also observed
that the average EC concentration were 2.34 (0.52), 3.71 (0.42), 4.99 (0.58), and 2.16
(0.47) Summer, Autumn, Winter and Spring at urban area in Xiamen, China. The average
EC concentrations in this study are relatively lower when compared with above studies for
distinct seasons.
The average WINSOC concentrations and associated standard deviation were 5.02
(0.45), 6.37 (2.36), 0.87 (0.24), 1.83 (0.18), 3.53 (0.30), and 4.18 (0.31) µg C m –3
for the
Autumn 2009, Winter 2010, Spring 2010, Summer 2010, Autumn 2010, and Winter 2011
seasons, respectively. During the 1st week of Spring 2011 (AVE 66) the average WINSOC
concentration and standard deviation was 1.61 (0.10) µg C m–3
. As it can be observed,
from Fig. IV-9, the levels of WINSOC tend to be higher in the colder seasons and lower in
the warmer periods. A similar seasonal trend in the average mass concentrations of
WINSOC was found by Kim et al. (2011) for the fine aerosol samples at an urban area.
During sampling campaign II, the WINSOC accounts for a significant fraction of OC,
ranging from 60 to 71, 63 to 78, 19 to 69, 24 to 82, 41 to 78, and 40 to 76% of the OC
content in Autumn 2009, Winter 2010, Spring 2010, Summer 2010, Autumn 2010 and
Winter 2010 seasons, respectively. Comparing the average WINSOC/OC ratio obtained in
this study for Summer season (55%) and Winter (61%), with those reported in the
literature, it is possible to verify that the average ratio obtained in Aveiro is higher than
those obtained in Atlanta (35%) during the summer season and lower than those obtained
in St. Louis (64%) during the Winter season (Sullivan and Weber, 2006). It has been
suggested that in urban areas, the WINSOC fraction is associated with anthropogenic
emissions (Miyazaki et al., 2006; Park and Cho, 2011), which comprises great amounts of
incompletely combusted products and biogenic detritus, such as wax esters, aliphatic
hydrocarbons, triglycerides, long-chain ketones, alkanols, and polycyclic aromatic
Global carbon balance and isolation of water-soluble organic matter from atmospheric aerosols
75
hydrocarbons (Mayol-Bracero et al., 2002). Similar indication was reported by Sun et al.
(2011), who found that WINSOC was associated with primary combustion emissions,
showing strong correlation with EC (r = 0.82), CO (r = 0.75), and NOx (r = 0.68). In fine
marine aerosol, Facchini et al. (2008) found that the WINSOC fraction was the most
important constituent of the TC component (average ratio and associated standard
deviation of 94 ± 4% of TC).
In what concerns the WSOC concentration, it accounts on average to 35, 27, 43, 33,
30, and 38 % of TC content in Autumn 2009, Winter 2010, Spring 2010, Summer 2010,
Autumn 2010, and Winter 2011 seasons, respectively. For the single aerosol sample
collected in Spring 2011, the WSOC/TC was about 41 %. The WSOC concentrations
during the sampling campaign II ranged between 0.368 (0.028) - 4.95 (0.13) µg C m-3
. A
large variability of WSOC concentration (0.821 (0.012)) - 4.95 (0.13) µg C m-3
) among
samples of Winter 2011, suggested much dependence or great influence of local
conditions. The concentrations of WSOC were found to be lower in Spring 2010 and
Summer 2010 seasons (average and standard deviation values of 1.05 (0.16) and 1.07
(0.15) µg C m-3
, respectively) and higher in Autumn 2009 and Winter 2011 seasons
(average and standard deviation values of 2.68 (0.23) and 2.57 (0.24) µg C m-3
,
respectively), as can be observed in Fig. IV-9. This seasonal variability in WSOC
concentrations in PM2.5 samples has been also reported in other studies performed at a
urban area (Jaffrezo et al., 2005; Viana et al., 2006b), a suburban location of India (Khare
et al., 2011), at a rural area of Portugal (Duarte et al., 2007). However, in a marine
location, the WSOC concentrations in aerosol samples collected during the Summer season
were found to be higher than those determined in samples collected during the Winter
season (Yoon et al., 2007). The authors suggested that this seasonal pattern is likely to be
associated to differences in biological activity between such seasons. The increase in the
domestic heating and the lower ambient temperatures which promote the particulate phase
in the gas-particle equilibrium (the SVOCs tend to be in the particulate phase at lower
temperatures (Kiss et al., 2002)). Overall, the results reported in the literature alongside
those described in the present study suggest that the WSOC component has different
primary sources (either anthropogenic and natural) and is dependent on the sampling area
and season. It was also observed an increase in the ambient concentrations of WSOC in
Chapter IV
76
weeks AVE 33 to 35 during the Summer 2010 season, coincident with the intense forest
fires events in the region of Aveiro.
Fig. IV-10 depict the annual variation of the WSOC/OC and OC/EC ratios for the
collected PM2.5 samples. These ratios have been used to deduce about SOA formation in
urban and rural sites (Ho et al., 2002; Viana et al., 2007; Ram and Sarin, 2010).
Fig. IV-10. Seasonal distribution of WSOC/OC and OC/EC concentration ratio in PM2.5 fraction during the
sampling campaign II.
Overall, the WSOC/OC ratios ranged from 22 to 81%, while the OC/EC ratios
varied between 1.3 and 16. Despite the wide variability, these ratios tend to be higher in
Spring and Summer seasons than in the other periods. The median WSOC/OC ratios were
36.8, 27.4, 56.4, 44.8, 37.1, and 37.6% for the Autumn 2009, Winter 2010, Spring 2010,
Summer 2010, Autumn 2010, and Winter 2011 seasons, respectively. The PM2.5 sample
collected in Spring 2011 exhibited a WSOC/OC ratio of 42.4%. The high median
WSOC/OC ratios observed during Spring and Summer seasons at the urban location
focused in this study suggest that the warmer ambient temperature, higher solar radiation,
and the likely presence of oxidants (e.g. O3 and OH) in the atmosphere may enhance the
oxidation of the particulate organic matter, leading to a higher water-solubility of the
organic fraction. Viana et al. (2007) have related the WSOC/OC ratios between 33 and
Weeks/Samples
AV
E 1
AV
E 2
AV
E 3
AV
E 4
AV
E 5
AV
E 6
AV
E 7
AV
E 8
AV
E 9
AV
E 1
0A
VE
11
AV
E 1
2A
VE
13
AV
E 1
4A
VE
15
AV
E 1
6A
VE
17
AV
E 1
8A
VE
19
AV
E 2
0A
VE
21
AV
E 2
2A
VE
23
AV
E 2
4A
VE
25
AV
E 2
6A
VE
27
AV
E 2
8A
VE
29
AV
E 3
0A
VE
31
AV
E 3
2A
VE
33
AV
E 3
4A
VE
35
AV
E 3
6A
VE
37
AV
E 3
8A
VE
39
AV
E 4
0A
VE
41
AV
E 4
2A
VE
43
AV
E 4
4A
VE
45
AV
E 4
6
AV
E 4
7
AV
E 4
8
AV
E 4
9
AV
E 5
0A
VE
51
AV
E 5
2A
VE
53
AV
E 5
4A
VE
55
AV
E 5
6A
VE
57
AV
E 5
8A
VE
59
AV
E 6
0A
VE
61
AV
E 6
2A
VE
63
AV
E 6
4A
VE
65
AV
E 6
6
WS
OC
/OC
rat
io (
%)
0
20
40
60
80
100
OC
/EC
rat
io
0
5
10
15
20
WSOC/OC (%)
OC/EC
2
Autu
mn 2
00
9
Winter 2010 Spring 2010 Summer 2010 Autumn 2010 Winter 2011
Spri
ng 2
01
1
Global carbon balance and isolation of water-soluble organic matter from atmospheric aerosols
77
43% in summer to the photochemical oxidation of primary organic aerosols, which
increased the WSOC fraction due to the formation of more oxygenated functional groups.
Average WSOC/OC ratios ranging from 34 to 61 during the warmer period (i.e., summer)
at different urban areas were reported by Sullivan and Weber (2006), Miyazaki et al.
(2006), and by Viana et al. (2007). Kondo et al. (2007) found that 35% (in median) of OC
was water-soluble at an urban location Tokyo, Japan, during summer. Furthermore, the
WSOC/OC ratios obtained in a previous study at a rural site near the city of Aveiro (Duarte
et al., 2007) were in the range of those found in this study. On the other hand, the
WSOC/OC ratios for samples collected between weeks AVE 33 and AVE 35 are higher
than those found in the final weeks of summer 2010 (namely week AVE 38 and AVE 39),
and comparable to the previous weeks, may suggest the influence of the forest fire events
occurred in the region of Aveiro during these sampling periods. Similarly, low median
WSOC/OC ratios (0.35-0.37) were found for PM2.5 samples collected during biomass
burning events in northern India (Ram and Sarin, 2010).
The OC/EC ratios (Fig. IV-10) ranged between 1.3 and 16 from week AVE 14 to
week AVE 50. For the aerosol samples collected in Autumn 2009, Winter 2010 (namely,
weeks AVE 7 to AVE 10), Winter 2011, and Spring 2011, the values of OC/EC (data not
shown) were in the range of 202.5- 430.3, 259.6- 405.3, 25.0-77.4, and 22.8, respectively.
These unusual high OC/EC ratios are a consequence of the very low EC concentrations
found in the aerosol samples, although the sampling site is under the influence of well
identified anthropogenic sources of EC (e.g., emissions from vehicle traffic, industrial, and
other human activities). Generally, OC/EC ratios above 2, except for samples AVE 25 and
AVE 31, can be indicative of the presence of secondary OC in atmospheric aerosols
(Castro et al., 1999). The lowest values of the OC/EC ratios were obtained for samples
collected in weeks AVE 25 and AVE 31 (1.4 and 1.3, respectively), suggesting that the OC
measured in these samples may have a primary origin. During these two sampling periods,
the sampling site was under the influence of air masses travelling from the continent
(Annex C) and, therefore, more enriched in anthropogenic EC emissions from the
surrounding urban areas. The low OC/EC ratios obtained for samples AVE 25 and AVE 31
(in Spring 2010 and summer 2010 seasons, respectively) may also be due to a low OC
concentration since during those periods the air temperature is higher, thus favoring the
gaseous phase in the gas-particle partitioning of the SVOCs. Furthermore, the OC/EC
Chapter IV
78
ratios obtained in this study, for the period in question (AVE 25 and AVE 31), are higher
than those obtained in other urban European locations (Castro et al., 1999). As shown in
Fig. IV-10, samples AVE 14 and AVE 49 exhibit the highest OC/EC ratios (10 and 16,
respectively). Note that for sample AVE 49, the concentration of OC was one of the
highest (Fig. IV-10), contributing to about 94% of the TC in fine atmospheric particles.
During the high period of forest fires (Summer B 2010) the OC/EC ratios ranged between
3.0 and 6.2. Saarikoski et al. (2008) have reported distinct OC/EC ratios values on
measurements at urban environment: 0.71 for traffic emissions; 3.3 for secondary organic
carbon; 6.6 for biomass combustion; and 12 for long-range transport. A large variability of
OC/EC ratios (range: 2.4–14.5, Av=7.87±2.4, n=77) were obtained at urban sites, which
can be related to a predominance of contributions from biomass burning sources (wood-
fuel and agriculture waste) (Ram and Sarin, 2010).
4.5. Seasonal sample division and natural event identification
The PM2.5 samples collected in sampling campaign II were grouped together
according to similar meteorological conditions (section 4.2, Fig. IV-1), air masses
trajectories (section 4.2, Fig. IV-3), and WSOC concentrations, on a total of 10 groups,
corresponding to different seasonal events, as shown in Table IV-3. Thus, in Autumn 2009,
samples collected in weeks AVE 1 to AVE 4 were grouped together. In Winter 2010,
because of the mentioned drawback during the sampling campaign (section 3.3, Chapter
3), sample collected in week AVE 14 was grouped with samples collected during weeks
AVE 16 and AVE 17 in Spring 2010; for this reason, this group was termed Winter/Spring
2010. In Spring 2010, it was possible to set two additional groups of samples, named
Spring A 2010 and Spring B 2010, where the samples of the former group were collected
during the volcanic eruption events occurred in Iceland. In Summer 2010, two different
natural events were recorded during the sampling campaign, thus originating two groups of
samples: Summer A 2010, where it was observed the intrusion of Sahara Desert dust over
the Iberian Peninsula, and Summer B 2010, where it was observed the occurrence of forest
fires episodes.
Global carbon balance and isolation of water-soluble organic matter from atmospheric aerosols
79
Table IV-3. Summarized information regarding the assembled PM2.5 samples collected in sampling campaign II.
Sample / Season Week Origin of air masses Natural
events
Median
Temperature
(°C)
Maximum
RH (%)
Maximum
Wind velocity
(m/s)
Accumulated
Rainfall
(mmH2O)
WSOC
(µg C m-3
)
Avg sd
Autumn 2009
AVE 1 Maritime
12.1 94 9.5 49.1 2.91 0.12
AVE 2 13.7 95 6.9 83.4 1.91 0.07
AVE 3 Maritime/Continental
12.6 92 7.5 0.1 3.01 0.16
AVE 4 8.1 91 7.5 20 2.90 0.10
Winter/Spring 2010
AVE 14 Continental
9.3 81 7.3 0.6 2.38 0.06
AVE 16 Maritime/Continental
12.5 90 7.7 8.6 1.76 0.05
AVE 17 15.8 92 6.4 1.8 1.63 0.05
Spring A 2010
AVE 19
Maritime/Continental Volcanic eruptions in
Iceland
17.1 92 5.9 52.5 0.89 0.00
AVE 20 16.4 89 11.3 0.5 1.05 0.05
AVE 21 14.7 92 7.4 11.4 0.83 0.06
AVE 22 14.2 93 10.2 2.9 0.89 0.04
Spring B 2010
AVE 23
Maritime/Continental
21.5 92 5.6 0.5 1.48 0.08
AVE 25 18.8 93 5.6 4.8 0.69 0.03
AVE 27 17.3 91 7.3 0 0.93 0.08
Summer A 2010
AVE 28 Continental
Sahara Desert dust
17.9 96 3.8 0 0.99 0.06
AVE 30 Maritime/Continental 18.6 93 5.3 0.1 0.67 0.01
AVE 31 Maritime 19.3 93 5.1 1.1 0.37 0.03
AVE 32 Maritime/Continental 19.6 93 7.1 0.2 0.62 0.02
Summer B 2010
AVE 33
Maritime/Continental Forest fires
20.8 93 4.8 0 1.83 0.08
AVE 34 20.2 90 6.2 0 1.36 0.09
AVE 35 19.6 95 6.3 0 2.26 0.05
Autumn A 2010
AVE 40 Maritime/Continental
18.7 94 6.4 1.6 1.34 0.08
AVE 41 Maritime
17.9 92 9.7 41.1 1.15 0.13
AVE 42 17.2 92 7.3 76.8 2.28 0.06
Autumn B 2010
AVE 46
Maritime/Continental
15 95 9.9 17.5 2.20 0.08
AVE 47 14.3 92 9.7 52 1.13 0.04
AVE 48 13.05 92 9.6 46.9 1.50 0.16
AVE 49 8.4 91 5.4 25 2.81 0.07
AVE 50 9.3 93 10.7 172.9 1.64 0.07
Winter 2011
AVE 54 Maritime/Continental
12.9 94 4.8 34 2.16 0.05
AVE 56 13.1 92 4.3 25.2 3.77 0.10
AVE 57 Continental
10.05 91 6.7 2.6 2.90 0.08
AVE 58 7.4 92 6.3 11.8 4.95 0.13
Winter/Spring 2011
AVE 63 Continental 11.65 76 7.3 0.1 2.89 0.09
AVE 64
Maritime/Continental
12.2 91 5.8 28 2.60 0.12
AVE 65 Pollen particles in
filters
13.3 94 6.5 0 2.19 0.05
AVE 66 13.8 94 7.5 38.1 1.18 0.09
Chapter IV
80
In Autumn 2010, two groups of samples were assembled, with the first group
(Autumn A 2010) being collected under ambient conditions more closely related to those
recorded in Summer 2010, and the second group (Autumn B 2010) being collected under
more colder temperatures and wet conditions than the previous group. In Winter 2011,
which was period with very unstable meteorological conditions, only one group of samples
was set. Finally, the group of samples named Winter/Spring 2011, include three samples
collected in the late Winter 2011 and one sample collected in early Spring 2011,
corresponding to a period of four consecutive weeks with similar ambient conditions.
Before the isolation and fractionation of the WSOM from the atmospheric particles,
the aqueous extracts from each quartz filter were grouped together according to the groups
of samples previously established. The reason for combining the aqueous extracts is to
ensure enough material for the subsequent structural characterization by the various
analytical techniques (Chapters 5 and 6), which means that the focus will be set on the
seasonal variability of the chemical characteristics of both PM2.5 and carbonaceous
material.
4.6. Impact of forest fire emissions on PM2.5 and carbonaceous material
Since forest fires are a major source of carbonaceous aerosol in many European
regions, it was decided to include in this study the evaluation of the impact on the
concentrations of both PM2.5 and carbonaceous material of the forest fire events that
occurred in the region of Aveiro. In order to accomplish this objective, four groups of
samples were selected: two samples collected in sampling campaign I (Summer 2008 and
Summer 2009) and two samples collected in sampling campaign II (Summer A 2010 and
Summer B 2010).
Over the past decade, forest fires have been recurrent and identified as a
problematic phenomenon mainly in the northern and center of Portugal. Fig. IV-11 shows
the number of forest fires events (“hot flashes” and fires) recorded during the selected
seasons. The higher temperatures during these summer periods (namely, in Summer A and
B 2010) alongside a RH around 20% (in minimum) helped to extend the duration and
Global carbon balance and isolation of water-soluble organic matter from atmospheric aerosols
81
spread of the forest fires. From Fig. IV-11, it can be noticed that the highest number of
forest fires events occurred in Summer B 2010 and the lowest number occurred in Summer
2009. The forest fires are responsible for the loss of many of the Portuguese forested areas,
and the report of Agriculture, Rural Development and Fisheries Ministry (MADRP-
Direcção de Unidade de Defesa da Floresta, 2011) reveals that the total burnt area of
forest land was about 17564 ha, 87420 ha, and 133090 ha in the years 2008, 2009, and
2010, respectively.
Fig. IV-11. Number of forest fires events recorded during the selected sampling seasons.
The concentrations of the total fine aerosol (PM2.5) and of the carbonaceous
materials (TC, OC, EC, and WSOC) for these periods are shown in Fig. IV-12. The PM2.5
mass concentrations varied from 8.2 to 13.5 gm-3
in Summer 2008 (SU08-1 – SU08-4),
from 8.9 to 10.5 gm-3
in Summer 2009 (SU09-1 – SU09-4), from 9.7 to 24.6 gm-3
in
Summer A 2010 (AVE 28, AVE 30–AVE 32), and from 20.5 to 27.0 gm
-3 in Summer B
2010 (AVE 33–AVE 35). The highest values of PM2.5 were registered in Summer B 2010
season, matching the period with the highest number of forest fire emissions and the
influence of air masses travelling from the continent surroundings. These results suggest
that the forest fire emissions may have an important contribution to the increase of the
Season
Summer 2008Summer 2009
Summer A 2010Summer B 2010
Num
ber
of o
ccur
renc
es
0
20
40
60
80
200
400
600
800
1000
1200
"Hot flashes"
Fires
Chapter IV
82
concentration of fine particles in the atmosphere. The lowest concentrations of fine air
particles were found during Summer 2008 and Summer 2009, when the sampling site was
under the influence of clean air masses travelling from the maritime surroundings.
Fig. IV-12. Ambient concentrations of PM2.5 (in µg m-3
) and TC, OC, EC, and WSOC (in µg C m-3
) for the
samples collected during Summer 2008, Summer 2009, Summer A 2010, and Summer B 2010.
Fig. IV-12 shows that the highest concentration values for OC and EC were
recorded during the Summer B 2010 season: 3.36 – 6.11 g C m-3
and 0.98 – 1.10 g C m-
3, respectively. During Summer 2008, Summer 2009, and Summer A 2010, the OC
concentrations varied between a minimum of 0.64 g C m-3
and a maximum of 1.35 g C
m-3
, whereas the EC concentrations ranged between 0.05 g C m-3
and 0.67 g C m-3
.
These levels of OC and EC are much lower than those reported for an urban site under
forest fires ambient (from 17 to 71 g C m-3
and between 8 and 31 g C m-3
, respectively)
(Mayol-Bracero et al., 2002), and also lower than those observed in Shanghai (from 1.1 to
2.7 g C m-3
) (Feng et al., 2006). For Summer 2008, Summer 2009, and Summer A 2010,
the concentrations of WSOC were found to be lower than 1 g C m-3
, whereas the highest
values were verified during summer B 2010 season (above 1.3 g C m-3
), as shown in Fig.
IV-12. The concentrations of WSOC during these latest seasonal period are comparable to
Samples
SU08-1
SU08-2
SU08-3
SU08-4
SU09-1
SU09-2
SU09-3
SU09-4
AV
E 28
AV
E 30
AV
E 31
AV
E 32
AV
E 33
AV
E 34
AV
E 35
PM
2.5
conc
entr
atio
n (
gm-3
)
0
5
10
15
20
25
30
Car
bon
conc
entr
atio
n (
gCm
-3)
0
1
2
3
4
5
6
7
8
PM2.5
TC
OC
EC
WSOC
Summer B 2010Summer A 2010Summer 2009Summer 2008
Global carbon balance and isolation of water-soluble organic matter from atmospheric aerosols
83
those reported by Feng et al. (2006) (0.7-1.9 g C m-3
) for a urban area during the summer
season; however, such values are much lower than those reported by Mayol-Bracero et al.
(2002) (11–46 g C m-3
), where the measurements of WSOC were made under highly
active forest fires events. In the present study, the lowest WSOC concentrations were
measured for the samples collected during Summer 2009 period, which is characterized by
air masses originating primarily from the maritime surroundings and by a low number of
forest fire events.
Fig. IV-13 depicts the WSOC/OC ratios for the aerosol samples collected during
the selected periods. The WSOC/OC ratios ranged from 40.8% to 60.3% in Summer 2008,
31.4% to 58.2% in Summer 2009, 47.8 % to 75.6 % in Summer A 2010, and 30.0% to 44.8
% in Summer B 2010. WSOC/OC ratios higher than 40% have been reported for aerosol
samples from biomass burning episodes (Mayol-Bracero et al., 2002; Saarikoski et al.,
2007). The ratios did not vary considerably throughout the summer seasons, but Summer
2010 A shows two weeks (AVE 28 and AVE 30) with higher WSOC/OC ratios. The low
WSOC/OC ratio for the SU09-4 and AVE 33 samples (31.4% and 30.0%, respectively) is
likely due to the low WSOC concentration in comparison to the levels of OC. This finding
suggests that OC collected during these weeks is less oxidized, and thus less water-soluble.
The low WSOC/OC ratios in the samples previously mentioned (SU09-4 and AVE 33)
may also reflect a larger contribution of local sources to the carbonaceous material, which
is also suggested by Miyazaki et al. (2006), who found a WSOC/OC ratio of 35% for fine
particles collected during summer 2004 in Tokyo.
Chapter IV
84
Fig. IV-13. WSOC/OC ratio for each PM2.5 sample collected during the warmer periods.
The OC/EC ratios ranged from 2.8 to 5.3, 7.2 to 21.0, 1.4 to 6.0 and 3.0 to 6.2 for
Summer 2008, Summer 2009, Summer A 2010, and Summer B 2010 seasons, respectively.
OC/EC ratios higher than 2.0 has been associated to the presence of secondary OC in
atmospheric particles (Chow et al., 1996; Castro et al., 1999; Krivacsy et al., 2001b; Ho et
al., 2003). The results previously reported suggest that in the small town of Aveiro during
the warmer periods, the OC and EC components have different emission sources, where
the increase in OC concentrations is not accompanied by significant increases also in EC.
Those results may also suggest the existence of a regional contribution of aged aerosol
with higher secondary OC content, thus increasing the OC/EC ratios. The OC/EC ratios
measured in the small town of Aveiro during the warmer periods were also found higher
than those reported for the more polluted area of Shanghai (average OC/EC ratio = 2.2)
(Feng et al., 2006), of Beijing (average OC/EC ratio = 2.4), and also during the Summer
season and without forest fire episodes ((Pio et al., 2008); OC/EC ratio < 3). For Summer
B 2010, corresponding to the period with the highest number of forest fire events, the
obtained OC/EC ratios (3.0-6.2) are of the same order of magnitude of those reported for
biomass burning aerosols (OC/EC ratios from 3 to 7) (Pio et al., 2008). These results
suggest that both biomass burning primary and secondary sources can jointly contribute to
the presence of OC in atmospheric particles in warmer periods.
Samples
SU08
-1
SU08
-2
SU08
-3
SU08
-4
SU09
-1
SU09
-2
SU09
-3
SU09
-4
AV
E 28
AV
E 30
AV
E 31
AV
E 32
AV
E 33
AV
E 34
AV
E 35
WSO
C/O
C (%
)
0
20
40
60
80
Summer B 2010Summer A 2010Summer 2009Summer 2008
Global carbon balance and isolation of water-soluble organic matter from atmospheric aerosols
85
4.7. Isolation and fractionation of water-soluble organic matter from
atmospheric particles
As previously mentioned, the isolation and fractionation of WSOM was performed
following the procedure of Duarte and Duarte (2005) and two fractions were obtained:
WSOC hydrophobic acids (recovered from the DAX-8 resin) and WSOC hydrophilic acids
(not retained in the resin).
4.7.1. Preliminary tests to improve the recovery of organic matter from the DAX-8
resin
During the isolation and fractionation procedure of each sample, the UV
absorbance at 250nm (UV250) and the DOC content of the each obtained fraction was
measured in order to account for the total amount of recovery and losses of chromophoric
organic compounds and DOC, respectively.
As mentioned in section 3.7, Duarte and Duarte (2005) employed a solution of
MeOH:H2O in the proportion 1:1 for recovering the organic matter adsorbed onto a XAD-
8 resin. In the present study, a DAX-8 resin was used instead and it was concluded that a
solution with a higher content of organic solvent (i.e. MeOH) is required for improving the
amount of organic matter (in terms of UV250 and DOC) that is actually recovered from the
DAX-8 resin. As such, two different solutions with different proportions of MeOH:H2O
(1:1 and 3:2) were tested using a Pony Lake fulvic acids reference sample (1R109F)
obtained from the IHSS. The distribution and average percentages of retention, recovery,
and loss of UV250 and DOC from three replicas of the isolation/fractionation procedure of
aqueous solutions of Pony Lake fulvic acids using different proportions of MeOH:H2O
(1:1 and 3:2) for recovering the adsorbed organic matter are shown in Table IV-4.
Table IV-4. Distribution and average percentages (± standard deviation) of retention, recovery, and loss of
UV250 and DOC from three replicas of the isolation/fractionation procedure of aqueous solutions
of Pony Lake fulvic acids. Three replicas were performed.
Chapter IV
86
Eluent
MeOH:H2O 1:1 MeOH:H2O 3:2
UV250 DOC UV250 DOC
Retained onto DAX-8
resin 83.0 (2.9) 73.4 (6.8) 83.0 (2.7) 77.2 (1.8)
Recovered from DAX-8
resin 41.6 (2.6) 31.5 (7.4) 62.8 (5.5) 54.7 (7.1)
Losses due to the
desalination procedure 6.5 (0.7) 3.0 (1.8) 7.7 (5.4) 3.8 (1.3)
Losses due to
irreversible adsorption
onto DAX-8 resin
34.9 (3.6) 38.8 (5.3) 12.5 (5.9) 18.8 (8.1)
Not retained in DAX-8
resin 17.0 (2.9) 26.6 (6.8) 17.0 (2.7) 22.8 (1.8)
As shown in Table IV-4, approximately 32% of DOC was recovered from the
DAX-8 resin using a 1:1 MeOH:H2O solution, whereas this percentage increases to almost
55% using a 3:2 MeOH:H2O solution. It can also be observed that approximately 39% and
19 % of DOC was irreversibly adsorbed onto the resin’s polymers when using solutions of
MeOH:H2O in proportions 1:1 and 3:2, respectively. Overall, an increase of approximately
23% in the amount of DOC recovered from the DAX-8 resin was achieved using a 3:2
MeOH:H2O solution. This solution also allowed an increase in the amount of UV250 that is
recovered from the resin of approximately 21%. The EEM fluorescence spectra of the
recovered fractions (data not shown) suggested that no significant effects on the
fluorescence features of the sample was verified using a 3:2 MeOH:H2O solution for the
elution step. Thus, in the isolation/fractionation procedure of the WSOM samples obtained
in sampling campaign II, it was decided to apply as eluent a 3:2 MeOH:H2O solution.
Table IV-5 summarizes the percentages of losses and recovery of DOC and UV250
during the isolation procedure. The percentages of UV250 and DOC retained in the DAX-8
resin were in the range of 65 to 90% and 51 to 70%, respectively. In terms of percentage of
UV250 recovered from the DAX-8 resin (here operationally designated as “WSOC
hydrophobic acids” fraction), it ranged from 51 to 68 %, while in terms of DOC it was
possible to recover between 42 and 53 %, with the lowest values being found for the
Summer A 2010 sample. Additionally, between 12 and 35 % of UV250 and between 30 and
49 % of DOC was not retained onto the DAX-8 resin.
Global carbon balance and isolation of water-soluble organic matter from atmospheric aerosols
87
Table IV-5. Distribution and percentage (average (± sd)) of recovery and loss of UV-Vis absorbance at 250 nm (UV250) and DOC from the isolation/fractionation
procedure of WSOM from atmospheric samples collected in sampling campaign II. (1number of replicate=2;
2number of replicate=3)
Sample/Season
Retained onto DAX-8
resin
Recovered from DAX-8
resin
Not retained in DAX-8
resin
Losses due to irreversible
adsorption onto DAX-8
resin
Losses due to the
desalination
UV250 DOC UV250 DOC UV250 DOC UV250 DOC UV250 DOC
Autumn 20091 88.4 (± 0.7) 65.7 (± 0.5) 64.3 (± 5.2) 50.9 (± 4.8) 11.6 (± 0.7) 34.3 (± 0.5) 21.6 (± 4.4) 10.8 (± 4.2) 2.5 (± 0.0) 3.9 (± 0.1)
Winter/Spring
20101 87.1 (± 2.6) 70. 0(± 2.4) 68.2 (± 2.2) 53.1 (± 2.7) 12.9 (± 2.6) 30.0 (± 2.4) 15.2 (± 0.5) 12.5 (± 1.6) 3.7 (± 0.9) 4.4 (± 1.3)
Spring A 20101 80.0 (± 1.0) 66.4 (± 3.9) 61.7 (± 4.9) 53.0 (± 0.9) 20.0 (± 1.0) 33.6 (± 3.9) 14.9 (± 3.7) 8.2 (± 3.2) 3.4 (± 0.2) 5.2 (± 0.1)
Spring B 20101 75.4 (± 2.5) 65.6 (± 2.9) 59.4 (± 1.3) 50.5 (± 2.1) 24.6 (± 2.5) 34.4 (± 2.9) 11.7 (± 0.2) 9.9 (± 0.8) 4.2 (± 0.9) 5.2 (± 0.1)
Summer A
20101 64.7 (± 1.5) 51.3 (± 4.3) 51.0 (± 2.0) 41.6 (± 2.1) 35.3 (± 1.5) 48.7 (± 4.3) 9.2 (± 0.4) 4.2 (± 2.8) 4.4 (± 0.1) 5.5 (± 0.7)
Summer B
20101 84.5 (± 1.7) 68.2 (± 0.2) 63.6 (± 6.3) 49.7 (± 6.2) 15.5 (± 1.7) 31.8 (± 0.2) 14.9 (± 3.0) 11.8 (± 5.3) 5.9 (± 1.7) 6.7 (± 0-6)
Autumn A
20101 82.8 (± 1.9) 70.5 (± 1.0) 66.0 (± 0.3) 51.0 (± 1.1) 17.2 (± 1.9) 29.5 (± 1.0) 13.1 (± 1.7) 14.0 (± 1.1) 3.7 (± 0.4) 5.5 (± 1.1)
Autumn B
20102 89.0 (± 0.7) 62.6 (± 2.2) 67.3 (± 4.3) 50.8 (± 2.3) 11.0 (± 0.7) 37.4 (± 2.2) 19.1 (± 4.7) 6.4 (± 0.6) 2.7 (± 0.1) 5.4 (± 0.2)
Winter 20112 90.5 (± 2.1) 60.1 (± 4.5) 60.0 (± 4.2) 44.9 (± 1.3) 9.5 (± 2.1) 39.9 (± 4.5) 27.6 (± 2.7) 8.0 (± 3.4) 2.9 (± 0.6) 7.2 (± 2.3)
Winter/Spring
20112 88.8 (± 2.0) 57.5 (± 2.2) 60.3 (± 2.2) 45.3 (± 1.7) 11.2 (± 2.0) 42.5 (± 2.2) 25.1 (± 1.6) 4.7 (± 1.6) 3.4 (± 0.7) 7.5 (0.9)
Chapter IV
88
It also can be observed in Table IV-5 that the values of the percentage of recovery
in terms of UV250 are higher than those in terms of DOC content. This difference was
explained by Duarte and Duarte (2005) as being due to the enrichment of the DAX-8
fraction in organic compounds with highly conjugated π bond systems. As already stated,
during the isolation and fractionation procedure, two kinds of losses must be accounted: (1)
losses due to the irreversible adsorption of organic compounds onto the resin polymers,
ranging from 4.2 to 14% in terms of DOC and 9.2 to 27.6% in terms of UV250; and (2)
losses due to the washout from the resin of the very weakly retained compounds that are
desorbed together with the inorganics during the desalting procedure, accounting for 3.9–
7.5% of DOC and 2.5–5.9% of UV250.
Table IV-6 shows a comparison of the percentages of retention and recovery of
DOC and UV250 obtained from the application of the isolation/fractionation procedure to
two aerosol WSOM samples collected at different locations: the Autumn 2009, Autumn A
2010, and Autumn B 2010 samples collected in this study at the urban location and an
Autumn sample collected by Duarte and Duarte (2005) at a rural location.
Table IV-6. Distribution and percentage (average (± sd)) retention and recovery of UV absorbance at 250 nm
(UV250) and DOC from the isolation/fractionation procedure of WSOM from atmospheric
samples collected in the Autumn season at two different locations.
Fraction
WSOM from
urban atmospheric
aerosols - Autumn
(2009) (this study)
WSOM from urban
atmospheric
aerosols - Autumn
A 2010 (this study)
WSOM from urban
atmospheric
aerosols - Autumn
B 2010 (this study)
WSOM from rural
atmospheric aerosols
– Autumn (Duarte
and Duarte, 2005)
DOC UV250 DOC UV250 DOC UV250 DOC UV250
Retained
onto the
resin
65.7
(± 0.5)
88.4
(± 0.7)
70.5
(± 1.0)
82.8
(± 1.9)
62.6
(± 2.2)
89.0
(± 0.7)
75.9
(±3.8)
93.6
(±2.2)
Recovery
from resin
50.9
(± 4.8)
64.3
(± 5.2)
51.0
(± 1.1)
66.0
(± 0.3)
50.8
(± 2.3)
67.3
(± 4.3)
54.1
(±7.1)
79.3
(±6.7)
As shown in Table IV-6 the DAX-8 resin allows the recovery of a slightly smaller
amount of WSOC hydrophobic acids ( 51% of DOC for all Autumns samples –this study)
as compared to that of the XAD-8 resin ( 54% of DOC) reported by Duarte and Duarte
(2005). Given that the percentages of UV250 and DOC retention and recovery for the XAD-
Global carbon balance and isolation of water-soluble organic matter from atmospheric aerosols
89
8 resin were higher than those for the DAX-8 resin, one can suggest that these differences
are likely to be related to the different physical characteristics (e.g. pore size) of the resins.
Using aquatic humic solutes, Peuravuori et al. (2002b) has reported that the sorptive power
of the DAX-8 resin was systematically somewhat greater compared to that of the XAD-8
resin. Nevertheless, the authors also concluded that the DAX-8 and XAD-8 resins seem to
isolate humic solute bulks almost equally, although the content of aliphatics is slightly
greater for the former, producing mixtures with similar structural compositions for general
purposes. In the present study, another relevant reason for the different sorptive power
exhibited by the two resins for the aerosol WSOM samples may be related with the
different nature of the WSOM samples since they were collected at two completely
different areas (urban and rural).
The average ambient concentrations (and standard deviations) of WSOC
hydrophilic and WSOC hydrophobic acids fractions (measured as DOC) obtained from the
isolation procedure of each group of aerosol samples representative of different seasonal
periods was also calculated and are summarized in Table IV-7. This table also presents the
amount (in mg) of solid residue of each WSOC hydrophobic acids fraction obtained at the
end of the freeze-drying procedure. As can be observed from Table IV-7, the Summer A
2010 sample contains the lowest amounts of WSOC hydrophobic acids. Lower levels of
WSOC hydrophobic acid were also reported by Sannigrahi et al. (2006), and Sullivan and
Weber (2006), for fine aerosols collected at an Atlanta urban site during stagnation events
during the summer season, when the air quality was poorest. Since there was little biomass
burning influence, Sullivan and Weber (2006) also concluded that SOA results from the
production of low molecular weight hydrophilic compounds.
As it can be seen from Table IV-7, the total amount of WSOC hydrophilic acids
fractions obtained for the various samples are typically lower than those of the WSOC
hydrophobic acids fractions, thus hindering the structural characterization of the
hydrophilic fractions. Therefore, the structural characterization that will be presented and
discussed in the next section is focused only on the WSOC hydrophobic acids fractions of
each group of aerosol samples.
Chapter IV
90
Table IV-7. Average (Avg.) ambient concentrations (in µg C m-3
) and associated standard deviation (sd) of
total WSOC, and isolated WSOC hydrophobic acids and WSOC hydrophilic acids fractions. The
amount (in mg) of solid residue of each WSOC hydrophobic acids fraction after freeze-drying is
also presented.
Season Filter/sample
WSOC
WSOC
Hydrophobic
Acid
WSOC
Hydrophilic
Acid
Amount of
solid residue
of each
WSOC
hydrophobic
acids
fraction
(µg C m-3
) (µg C m-3
) (µg C m-3
) (mg)
Avg. sd Avg. sd Avg. sd
Autumn 2009
AVE 1
2.68 0.23 1.36 0.13 0.92 0.01 50.6 AVE 2
AVE 3
AVE 4
Winter/Spring 2010
AVE 14
1.93 0.09 1.02 0.05 0.58 0.05 25.8 AVE 16
AVE 17
Spring A 2010
AVE 19
0.91 0.09 0.48 0.01 0.31 0.04 13.9 AVE 20
AVE 21
AVE 22
Spring B 2010
AVE 23
0.94 0.12 0.48 0.02 0.32 0.03 19.6 AVE 25
AVE 27
Summer A 2010
AVE 28
0.66 0.07 0.28 0.01 0.32 0.03 15.0 AVE 30
AVE 31
AVE 32
Summer B 2010
AVE 33
1.82 0.13 0.90 0.11 0.58 0.00 35.7 AVE 34
AVE 35
Autumn A 2010
AVE 40
1.11 0.16 0.57 0.01 0.33 0.01 12.8 AVE 41
AVE 42
Autumn B 2010
AVE 46
1.86 0.21 0.96 0.03 0.67 0.01 52.9
AVE 47
AVE 48
AVE 49
AVE 50
Winter 2011
AVE 54
3.44 0.19 1.57 0.02 1.46 0.02 60.7 AVE 56
AVE 57
AVE 58
Winter/Spring 2011
AVE 63
2.22 0.18 0.99 0.03 0.96 0.05 43.8 AVE 64
AVE 65
AVE 66
Global carbon balance and isolation of water-soluble organic matter from atmospheric aerosols
91
4.8. Aerosol mass balance of fine atmospheric aerosols
An aerosol mass balance was performed for assessing the contribution of each
carbonaceous fraction (namely, WSOC, WINSOC, and EC) for the total PM2.5 aerosol
mass. The relative contribution of each carbonaceous fraction to the total PM2.5 mass
during sampling campaign II at the city of Aveiro is shown in Fig. IV-14.
The mass of WSOM, water-insoluble organic matter (WINSOM), and EC fractions
were estimated by applying a conversion factor to the concentrations of WSOC, WINSOC
and EC concentration, respectively. In this work, the conversion factor employed for
estimating the amount of WSOM was 1.6, which was determined based on the results of
elemental analysis (see section 5.3) of the WSOM samples. A similar value has been
suggested in the literature for estimating the amount of WSOM in atmospheric aerosols
collected in urban areas (Turpin and Lim, 2001; Timonen et al., 2010). For the estimative
of the amount of WINSOM and EC, it was decided to use a conversion factor of 1.2 and 1,
respectively, in a similar fashion to what has been applied in previous studies (Zappoli et
al., 1999; Krivacsy et al., 2001b; Duarte, 2006).
As shown in Fig. IV-14, the total fine aerosol mass contains a highly variable
fraction (between 16 and 47%) of carbonaceous organic matter, while 53 to 84% remains
to be identified (NID). This component probably comprises inorganic species and trace
metals as suggested by Duarte (2006). For marine aerosols, Claeys et al. (2010) suggested
that the crustal matter (mineral dust) could be a major aerosol component of the NID
fraction. The carbonaceous fraction in PM2.5 tends to be higher during Autumn 2009 and
Winter 2011 (47%) periods, whereas the lowest levels were found for Spring B 2010 and
Summer A 2010 seasons (~16%). The EC fraction accounted for less than 6% of the total
PM2.5 mass during the annual sampling campaign II. The amount of WSOM was higher
than that of WINSOM for samples collected in Winter/Spring 2010, Spring A 2010, Spring
B 2010, Summer A 2010, Autumn A 2010, and Winter/Spring 2011 seasons and the
opposite in the other samples.
Chapter IV
92
Fig. IV-14. Aerosol mass balance for the fine atmospheric particles collected in Aveiro during Autumn 2009, Winter/Spring 2010, Spring A 2010, Spring B 2010,
Summer A 2010 and Summer B 2010, Autumn A 2010, Autumn B 2010, Winter 2011, and Winter/Spring 2011 seasons. "NID" refers to the mass of
aerosol that was not identified.
Global carbon balance and isolation of water-soluble organic matter from atmospheric aerosols
93
Fig. IV-14. Continued.
Chapter IV
94
On the other hand, the Summer B 2010 sample exhibit a higher percentage of
WINSOM as compared to that of the WSOM, suggesting a higher contribution of less
water-soluble primary organic aerosols, which seems to be in agreement with the
occurrence of local forest fires emissions. In a previous study conducted by Duarte (2006),
it was reported that the WSOM accounted for 30%, WINSOM for 13%, and EC for 2% of
the total fine aerosol mass collected during the Winter season at the city of Aveiro. The
unidentified fraction accounted for 55% of the total fine aerosol mass. In the present study,
for the Winter 2011 season, it was found that both WSOM and WINSOM fractions
account for 23%, EC for 1%, and NID for 53% of the total fine aerosol mass. The
difference between the values reported by Duarte (2006) and those obtained in this study
for the amount of WSOM and WINSOM may be related to the fact that the sampling time
applied by Duarte (2006) was only one week while in this study the values shown in Fig.
IV-14 correspond to a 4-week composite sample.
4.9. Conclusions
The assessment of the adsorption phenomena of SVOC and VOC during the
sampling procedure of atmospheric aerosols using quartz fibre filters revealed that these
phenomena accounts for less than 27% and 13% of the total PM2.5 and PM2.5-10 mass,
respectively. On the other hand, these phenomena can contribute to 22-33% and 4-7% of
OC present in PM2.5 and PM2.5-10 samples, respectively.
Overall, the seasonal trend found for the total PM2.5 mass concentrations was as
follows: Winter 2010>Winter 2011>Autumn2009>Autumn 2010>Summer 2010>Spring
2010. For the total PM2.5–10 mass concentrations, no clear seasonal trend can be foreseen;
however, the seasonal variability was as follows: Summer 2010>Winter 2010>Spring
2010>Autumn2009>Autumn 2010> Winter 2011. Episodes of high PM2.5 and PM2.5–10
mass concentrations in Summer B 2010 season are clearly associated with emissions from
local forest fires events. Carbonaceous materials (WSOM + WINSOM + EC) have an
important contribution to the total PM2.5 mass, accounting for 16-47% of the total fine
particulate matter. A considerable fraction (53-84%) of the total PM2.5 mass remains to be
identified, although it is likely that this component comprises inorganic species and trace
Global carbon balance and isolation of water-soluble organic matter from atmospheric aerosols
95
metals, as previously suggested by Duarte (2006). The WSOC is also preferentially
distributed in the fine aerosol samples, accounting for a highly variable fraction (27% to
41%) of the TC in PM2.5. The highest ambient WSOC concentrations were found in the fall
and Winter seasons, and the lowest values in the Spring and Summer seasons. The
WSOC/OC ratios are generally high (22-81%), which suggests that the contribution of
SOA is important for the total amount of WSOC in fine particulate matter.
The DAX-8 resin allowed the pre-concentration, isolation, and fractionation of the
organic matter dissolved in the aqueous extracts of the atmospheric aerosol samples. Two
fractions were obtained with this procedure: WSOC hydrophilic acids and WSOC
hydrophobic acids. The WSOC hydrophobic acids fractions, which were recovered from
the resin with a solution of MeOH:H2O in the proportion 3:2, represents 42 to 53% of the
total WSOC present in the aqueous extracts. Being the most important fraction, the
structural characterization that will be presented and discussed in the next chapter is
focused only on the WSOC hydrophobic acids fractions.
Chapter IV
96
97
V
Structural characterization of water-soluble
organic matter from fine urban atmospheric
aerosols
Chapter V
98
Structural characterization of water-soluble organic matter from fine urban atmospheric aerosols
99
5.1. Introduction
In this chapter, the results obtained from the application of Ultraviolet-Visible
(UV-Vis) and excitation-emission matrix (EEM) fluorescence spectroscopies to the WSOC
fractions of the various fine atmospheric aerosol samples are presented and discussed. It is
also presented and discussed the results of the structural characterization of the WSOC
hydrophobic acids fractions from the atmospheric aerosols collected in the different
seasons using the following techniques: Fourier transform infrared - attenuated total
reflectance (FTIR-ATR) and solid-state cross-polarization with magic angle spinning 13
C
nuclear magnetic resonance (CPMAS 13
C NMR) spectroscopies, and elemental analysis.
The study of the WSOM from atmospheric aerosols complemented with meteorological
data analysis will lead to a better understanding of the mechanisms of WSOM formation as
well as its sources for the city of Aveiro.
5.2. UV-Vis and EEM fluorescence spectroscopy of the aerosol water-
soluble organic carbon
5.2.1. UV-Vis spectroscopy
Before the isolation/fractionation procedure of the aerosol WSOM samples, the
UV-Vis spectra of all aqueous extracts were acquired. The typical UV-Vis spectra of
aerosol WSOC fractions are shown in Fig. V-1. In order to avoid concentration effects and
facilitate the comparison between the different fractions, the spectra are shown as specific
absorptivity (ɛ, g−1
C L cm−1
) versus wavelength (nm). These spectra are featureless and
characterized by a monotonically decrease of the specific absorptivity with increasing
wavelength.
Chapter V
100
Fig. V-1. UV–Vis spectra (specific absorptivity ɛ (g
−1 C L cm
−1) vs. wavelength (nm)) of the WSOC fractions extracted from the aerosol samples collected in
the different seasons.
Autumn 2009
(nm)
220 240 260 280 300 320 340 360 380 400 420 440 460 480 500
(g -
1 C
L c
m -
1)
0
10
20
30
40
50
60
AVE 1
AVE 2
AVE 3
AVE 4
Winter/Spring 2010
(nm)
220 240 260 280 300 320 340 360 380 400 420 440 460 480 500
(g -
1 C
L c
m -
1)
0
10
20
30
40
50
60
AVE 14
AVE 16
AVE 17
Spring A 2010
(nm)
220 240 260 280 300 320 340 360 380 400 420 440 460 480 500
(g -
1 C
L c
m -
1)
0
10
20
30
40
50
60
AVE 19
AVE 20
AVE 21
AVE 22
Spring B 2010
(nm)
220 240 260 280 300 320 340 360 380 400 420 440 460 480 500
(g -
1 C
L c
m -
1)
0
10
20
30
40
50
60
AVE 23
AVE 25
AVE 27
Summer A 2010
(nm)
220 240 260 280 300 320 340 360 380 400 420 440 460 480 500
(g -
1 C
L c
m -
1)
0
10
20
30
40
50
60
AVE 28
AVE 30
AVE 31
AVE 32
Summer B 2010
(nm)
220 240 260 280 300 320 340 360 380 400 420 440 460 480 500
(g -
1 C
L c
m -
1)
0
10
20
30
40
50
60
AVE 33
AVE 34
AVE 35
Structural characterization of water-soluble organic matter from fine urban atmospheric aerosols
101
Fig. V-1. Continued.
Autumn A 2010
(nm)
220 240 260 280 300 320 340 360 380 400 420 440 460 480 500
(g-1
C L
cm
-1)
0
10
20
30
40
50
60
AVE 40
AVE 41
AVE 42
Autumn B 2010
(nm)
220 240 260 280 300 320 340 360 380 400 420 440 460 480 500
e (
g-1
C L
cm
-1)
0
10
20
30
40
50
60
AVE 46
AVE 47
AVE 48
AVE 49
AVE 50
Winter 2011
(nm)
220 240 260 280 300 320 340 360 380 400 420 440 460 480 500
(g-1
C L
cm
-1)
0
10
20
30
40
50
60
AVE 54
AVE 56
AVE 57
AVE 58
Winter/Spring 2011
(nm)
220 240 260 280 300 320 340 360 380 400 420 440 460 480 500
(g-1
C L
cm
-1)
0
10
20
30
40
50
60
AVE 63
AVE 64
AVE 65
AVE 66
Chapter V
102
A similar behaviour has been reported in previous studies focusing on the
spectroscopic characterization of the WSOC fraction from atmospheric aerosols (Krivacsy
et al., 2001a, 2008; Duarte and Duarte, 2005; Duarte et al., 2005). Nevertheless, some UV-
Vis spectra exhibit an ill-defined shoulder between 250 and 300 nm, which has been
attributed to the presence of conjugated -bond systems (Duarte et al., 2003).
Little structural information can be drawn from the UV-Vis spectra of the aerosol
WSOC, but the examination of the values of the two different parameters, the quotient
E2/E3 (absorbances at 250 and 365 nm) and the specific absorptivity at 280 nm (ɛ280, g−1
C
L cm−1
), some qualitative information can be made. Fig. V-2 shows the obtained results for
both the ratio E2/E3 and ɛ280. Apparently, there is a seasonal trend in the values of both
parameters despite the similarity of the UV-Vis spectra. A clear seasonal variation of the
E2/E3 ratio can be observed, with the highest values being found for samples collected
during the warmer periods. This seasonal pattern of the E2/E3 ratios can be associated with
differences in the chemical structure of the aerosol WSOM typical of the different seasons;
usually, larger E2/E3 is representative of a shifting in UV–Vis absorbance towards low
wavelengths associated with a lower degree of humification of the WSOM (Duarte et al.,
2005). A similar seasonal trend was also verified by Duarte et al. (2005) at rural location,
and by Krivacsy et al. (2008) and Baduel et al., (2010) at an urban site. The variation in the
E2/E3 ratios observed between samples within the same season may be due to the different
nature of the WSOM in the PM2.5 sampled under different atmospheric conditions.
Structural characterization of water-soluble organic matter from fine urban atmospheric aerosols
103
Fig. V-2. E2/E3 ratio and ɛ280 (g−1
C L cm−1
) of the aerosol WSOC extracts collected at the different
seasons.
According to Peuravuori and Pihlaja (1997), there is also a relationship between the
ratio E2/E3 and the percentage of aromaticity and the molecular size of aquatic dissolved
organic matter. Higher E2/E3 ratios are generally associated with lower molecular sizes and
lower degree of aromaticity. Thus, a pattern such of that reported in Fig. V-2 suggests that
Season
Autum
n 2009
Win
ter/S
pring 2
010
Spring A
2010
Spring B
2010
Summ
er A
2010
Summ
er B
2010
Autum
n A 2
010
Autum
n B 2
010
Win
ter 2
011
Win
ter/S
pring 2
011
E2/E
3
3
4
5
6
7
8
9
10
Season
Autum
n 2009
Win
ter/S
pring 2
010
Spring A
2010
Spring B
2010
Summ
er A
2010
Summ
er B
2010
Autum
n A 2
010
Autum
n B 2
010
Win
ter 2
011
Win
ter/S
pring 2
011
28
0 (g
-1 C
L c
m -1
)
3
6
9
12
15
18
21
Chapter V
104
the organic matter of the water-soluble fraction of the aerosol samples collected in the
Summer A 2010 and Summer B 2010 seasons exhibit a lower molecular size and a lower
degree of aromaticity than those collected in the Autumn 2009, Winter/Spring 2010,
Autumn B 2010, Winter 2011, and Winter/Spring 2011 seasons.
The values of ɛ280 for the aerosol WSOC fractions (see Fig. V-2) also show that the
specific absorptivity increases for samples collected during the colder periods. Such
behavior suggests that these samples hold compounds with complex unsaturated bond
systems, where more than two -bond orbitals overlap leading to an increase in the
absorptivities values (Duarte et al., 2005). These results are in good agreement with those
derived from the E2/E3 ratio. The analysis of the specific absorptivity values in Fig. V-2
also reveals that the WSOC samples collected within the Summer B 2010 period, exhibit
higher ɛ280 values than those for Spring, Summer A 2010, and Autumn A 2010 seasons.
These results suggest the Summer B 2010 samples may be more enriched in unsaturated
bond systems, which could likely be a consequence of the impact of local forest fires
occurring in this period on the chemical characteristics of the WSOC.
5.2.2. EEM fluorescence spectroscopy
Although the EEM spectra of all aerosol WSOC fractions collected at the different
seasons have been recorded, it was decided to show in this section only some selected
examples of the acquired spectra, namely the EEM spectra of the WSOC fraction from the
first aerosol sample collected at each main season. The complete set of EMM spectra is
presented in Annex B. To avoid concentration effects and in order to facilitate the
qualitative comparison between the samples, the spectra were normalised for the DOC
content of the sample (in g L−1
) and to maximum fluorescence intensity and are shown as
specific fluorescence intensity (g−1
C L) versus excitation and emission wavelengths (nm).
The typical EEM spectra of the aerosol WSOC samples from the selected seasons (AVE 1
– Autumn 2009) are illustrated in Fig. V-3. These spectra depict four distinct fluorophores,
identified here by letters A, B, C, and D on the EEM spectrum of the AVE 1 – Autumn
Structural characterization of water-soluble organic matter from fine urban atmospheric aerosols
105
2009 sample: peak A at approximately λExcitation/λEmission (λEx/λEm) = 240/400-410, peak B at
λEx/λEm = 310-320/415-418, peak C at λEx/λEm = 280/340, and peak D at λEx/λEm = 230/340.
A similar fluorescence signature was observed for all the aerosol WSOC samples collected
in sampling campaign II. In general, fluorophores of type A and B are associated with the
presence of fulvic- and humic-like fluorophores (Coble, 1996), and the results obtained is
this study may indicate presence of constituents with similar fluorescence features in the
aerosol WSOM fractions. Bands in the same λEx/λEm range as those of A and B have also
been identified in EEM fluorescence spectra of both microbial and terrestrial derived
aquatic fulvic acids (McKnight et al., 2001). Peak C in marine samples has been assigned
to protein-like fluorophores (Coble, 1996; Yamashita and Tanoue, 2003), and to aromatic
amino acids (e.g. tryptophan) in freshwaters samples (Peuravuori et al., 2002a). In
atmospheric aerosols, peak C could also be due to aromatic nitrogen organic species
present in the collected atmospheric particles (Facchini et al., 1999). Further, other species
such as phenol-like moieties, e.g., catechol (Simoneit et al., 1999), 5-isopropyl-2-methyl-
phenol and 3-tert-butyl-phenol (Alves et al., 2001), may contribute to fluorescence of
WSOC at λEx/λEm = 280/340. A similar EEM fluorescence profile exhibiting bands at the
same λEx/λEm of peaks A, B, and C have already been identified in WSOC fractions from
atmospheric aerosols collected at a rural site (Duarte et al., 2004) and in DOM from
rainwater samples (Santos et al., 2009). Peak D at λEx/λEm = 230/340 has also been
assigned to aromatic amino acids, such as tyrosine, in marine samples (Coble, 1996;
Yamashita and Tanoue, 2003).
Chapter V
106
Fig. V-3. EEM spectra (specific fluorescence intensity (g−1
C L) versus excitation and emission wavelengths (nm)) of the WSOC fractions extracted from the
aerosol samples collected in Autumn 2009, Spring A 2010, Summer A 2010 and Winter 2011 seasons.
Structural characterization of water-soluble organic matter from fine urban atmospheric aerosols
107
According to Chen et al. (2002), synchronous fluorescence provides a better
sensitivity and peak resolution in comparison to the conventional emission fluorescence
scan mode, thus enabling the differentiation between the fluorescence spectra of samples
from different sources. The synchronous fluorescence spectra with a λ of 60 nm were
obtained from the EEM profiles of each WSOC fraction by fitting the mathematical
equation λEm =λEx + λ to the EEM profiles. The obtained spectra are shown in Fig. V-4
data were normalized to the DOC content (in g L−1
) of the sample. All synchronous
fluorescence spectra of the aerosol WSOC samples show a well-defined peak at a λEx of
approximately 330 nm, but with different specific fluorescence intensity values. The peak
at λEx 330 nm in sea water samples has been assigned to marine humic-like material
(Yamashita and Tanoue, 2003). The synchronous fluorescence spectra of the aerosol
WSOC samples from Autumn 2009, Autumn B 2010, and Winter 2011 seasons also
exhibit second well-defined band at approximately 280 nm, whose relative specific
fluorescence intensity is lower than that of peak at λExc 330 nm. Duarte et al. (2004)
suggested that the peak at λExc 280 nm could be due to aromatic nitrogen organic species
and phenol-like moieties present in the collected atmospheric particles. A similar
fluorescence pattern, but considerable more intense, has also been identified in humic and
fulvic acid fractions isolated from Kraft pulp mill effluents, being attributed to lignin-
derived structural moieties (Duarte et al., 2003). The fluorescence at λExc 280 nm in the
aerosol WSOC samples collected in the colder periods could also be due to the presence of
lignin-derived structures resulting from wood burning processes in domestic fireplaces for
house heating during those periods (Duarte, 2006). This band also appears in the
synchronous spectra of aerosol WSOC samples collected in warmer periods (Spring and
Summer), but as an ill-defined shoulder, being almost absent in the spectra of Summer A
2010 samples. Surprisingly, this band is more pronounced in the synchronous spectra of
the aerosol WSOC samples collected in Summer B 2010 period under intense forest fire
emissions, which supports the assumption that the direct emission into the atmosphere of
organic compounds from biomass burning processes play an important role in the bulk
chemical properties of WSOC fractions.
Chapter V
108
Fig. V-4. Synchronous fluorescence spectra with λ of 60 nm of the WSOC extracts from aerosol samples collected in Autumn 2009, Winter/Spring 2010,
Spring A 2010, Spring B 2010, Summer A 2010, Summer B 2010, Autumn A 2010, Autumn B 2010, Winter 2011, and Winter/Spring 2011 seasons.
Autumn 2009
Excitation (nm)
220 240 260 280 300 320 340 360 380 400Sp
ecif
ic F
luore
scen
ce I
nte
nsit
y (
g -
1 C
L)
0.0
0.1
0.2
0.3
0.4
0.5
0.6
AVE 1
AVE 2
AVE 3
AVE 4
Winter/Spring 2010
Excitation (nm)
220 240 260 280 300 320 340 360 380 400Sp
ecif
ic F
luore
scen
ce I
nte
nsit
y (
g -
1 C
L)
0.0
0.1
0.2
0.3
0.4
0.5
0.6
AVE 14
AVE 16
AVE 17
Spring A 2010
Excitation (nm)
220 240 260 280 300 320 340 360 380 400Sp
ecif
ic F
luore
scen
ce I
nte
nsit
y (
g -
1 C
L)
0.0
0.1
0.2
0.3
0.4
0.5
0.6
AVE 19
AVE 20
AVE 21
AVE 22
Spring B 2010
Excitation (nm)
220 240 260 280 300 320 340 360 380 400Sp
ecif
ic F
luore
scen
ce I
nte
nsit
y (
g -
1 C
L)
0.0
0.1
0.2
0.3
0.4
0.5
0.6
AVE 23
AVE 25
AVE 27
Summer A 2010
Excitation (nm)
220 240 260 280 300 320 340 360 380 400Sp
ecif
ic F
luore
scen
ce I
nte
nsit
y (
g -
1 C
L)
0.0
0.1
0.2
0.3
0.4
0.5
0.6
AVE 28
AVE 30
AVE 31
AVE 32
Summer B 2010
Excitation (nm)
220 240 260 280 300 320 340 360 380 400Sp
ecif
ic F
luore
scen
ce I
nte
nsit
y (
g -
1 C
L)
0.0
0.1
0.2
0.3
0.4
0.5
0.6
AVE 33
AVE 34
AVE 35
Structural characterization of water-soluble organic matter from fine urban atmospheric aerosols
109
Fig. V-4. Continued.
Autumn A 2010
Excitation (nm)
220 240 260 280 300 320 340 360 380 400
Sp
ecif
ic F
luore
scen
ce I
nte
nsit
y (
g-1
C L
)
0.0
0.1
0.2
0.3
0.4
0.5
0.6
AVE 40
AVE 41
AVE 42
Autumn B 2010
Excitation (nm)
220 240 260 280 300 320 340 360 380 400
Sp
ecif
ic F
luore
scen
ce I
nte
nsit
y (
g-1
C L
)
0.0
0.1
0.2
0.3
0.4
0.5
0.6
AVE 46
AVE 47
AVE 48
AVE 49
AVE 50
Winter 2011
Excitation (nm)
220 240 260 280 300 320 340 360 380 400
Sp
ecif
ic F
luore
scen
ce I
nte
nsit
y (
g-1
C L
)
0.0
0.1
0.2
0.3
0.4
0.5
0.6
AVE 53
AVE 54
AVE 56
AVE 57
AVE 58
Winter/Spring 2011
Excitation (nm)
220 240 260 280 300 320 340 360 380 400
Sp
ecif
ic F
luore
scen
ce I
nte
nsit
y (
g-1
C L
)0.0
0.1
0.2
0.3
0.4
0.5
0.6
AVE 63
AVE 64
AVE 65
AVE 66
Chapter V
110
5.3. Elemental analysis of aerosol water-soluble organic carbon
hydrophobic acids
The elemental analysis data (average values and associated standard deviation, on
an ash- and moisture free basis) and atomic ratios of the isolated WSOC hydrophobic acids
from the atmospheric aerosol samples collected at different seasons are reported in Table
V-1. The values obtained for the various WSOC hydrophobic acids fractions are of the
same order of magnitude and no distinct seasonal variation can be discern, although a
slight decrease in the relative quantity of carbon (%C) with concomitant increase in the
relative quantities of hydrogen (%H), nitrogen (%N), and oxygen (%O) can be observed
for the Spring and Summer samples.
Table V-1. Elemental composition (average (avg) and associated standard deviation (sd)) and atomic ratios
of aerosol WSOC hydrophobic acids collected at different seasons in sampling campaign II.
Three replicas were performed.
Sample % C % N % H % O Atomic ratios
avg sd avg sd avg sd avg sd H/C O/C N/C
Autumn 2009 65.7 0.06 2.64 0.06 5.83 0.05 25.9 0.16 1.07 0.31 0.034
Winter/Spring
2010 61.9 0.48 3.23 0.01 5.86 0.03 29.0 0.52 1.14 0.36 0.045
Spring A 2010 59.3 0.49 3.93 0.07 6.21 0.05 30.6 0.45 1.22 0.39 0.063
Spring B 2010 57.9 1.56 4.52 0.17 6.25 0.20 31.4 1.92 1.27 0.40 0.067
Summer A
2010 58.2 0.27 4.19 0.10 6.32 0.09 31.3 0.46 1.31 0.44 0.063
Summer B
2010 60.4 0.18 2.34 0.05 6.08 0.02 31.2 0.21 1.20 0.40 0.033
Autumn A
2010 61.6 0.71 3.66 0.06 6.27 0.09 28.5 0.82 1.20 0.35 0.050
Autumn B
2010 62.9 0.37 2.35 0.01 5.68 0.03 29.1 0.40 1.10 0.35 0.032
Winter 2011 63.7 0.19 2.37 0.03 5.61 0.05 28.4 0.18 1.06 0.35 0.032
Winter/Spring
2011 61.6 0.25 2.62 0.02 5.64 0.04 30.1 0.29 1.11 0.37 0.037
Overall, the values of C, H, N, and O were in the range of 57.9 (1.56)% (Spring B
2010) to 65.7 (0.06)% (Autumn 2009), 5.61 (0.05)% (Winter 2011) to 6.32 (0.09)%
(Summer A 2010), 2.34 (0.05)% (Summer B 2010) to 4.52 (0.17)% (Spring B 2010),
and 25.9 (0.16)% (Autumn 2009) to 31.4 (1.92)% (Spring B 2010), respectively. For
WSOC samples from fine atmospheric aerosols collected during the Summer period at the
Structural characterization of water-soluble organic matter from fine urban atmospheric aerosols
111
high-alpine site of Jungfraujoch, Switzerland, Krivacsy et al. (2001a) reported the
following elemental analysis data (average values and standard deviation in brackets) 52.3
(0.2)% for C, 6.7 (0.4)% for H, 2.5 (0.9)% for N, and 38.5 (1.1)% for O. Similar elemental
composition data were reported for a rural site in Hungary (K-puszta): 52% for C, 6.2% for
H, 2.5% for N, and 39% for O (Kiss et al., 2002). Duarte et al. (2007) also reported
elemental analysis data for WSOC samples from fine atmospheric aerosols collected
during one year period at a rural site near the city of Aveiro. The obtained values for C, H,
N, and O were in the range of 51.5-58.3%, 5.7-6.2%, 2.1-3.8%, and 31.8-36.6% ,
respectively, with no distinct seasonal variation. In comparison to the values reported in
the literature, the WSOC hydrophobic acids fractions at the city of Aveiro exhibit a higher
relative quantity of C and a lower relative quantity of O. Apparently, the WSOC samples
collected in this study are less oxidized, thus reflecting the important role played by local
anthropogenic sources on the structural features of urban aerosol WSOC as compared to
those collected at more clean environments (Krivacsy et al., 2001a; Kiss et al., 2002;
Duarte et al., 2007).
The examination of the atomic ratios (H/C, O/C, and N/C) also allow for some
qualitative estimation to be made. For sedimentary fulvic and humic acids from aquatic
and terrestrial environments, a high H/C atomic ratio has been related to a lower content of
unsaturated structures (Giovanela et al., 2004). On this regard, the WSOC hydrophobic
acids from Spring B 2010 and Summer A 2010 seasons show the highest H/C atomic
ratios, thus suggesting a higher aliphatic character of these samples compared to those of
the less warmer seasons. In comparison to the H/C atomic ratios reported by Duarte et al.
(2007) for fine aerosol WSOC samples collected at a rural site near the city of Aveiro, the
values obtained in present study are lower, thus suggesting a lower aliphatic character of
the urban aerosol WSOC samples (this feature is consistent with the results obtained in the
CPMAS 13
C NMR spectroscopy, in section 5.5).
The highest value for the O/C atomic ratio, which is used as an indicator of the
oxidation level of organic matter from terrestrial and aquatic environment (Abbt-Braun et
al., 2004), was obtained for the aerosol WSOC hydrophobic acids collected in Summer A
2010 season (as shown in Table V-1). Indeed, the O/C ratios also seem to follow a seasonal
trend, with the lowest values being determined for samples collected during the colder
Chapter V
112
periods. These findings suggest that the aerosol WSOC hydrophobic acids samples
collected during these periods exhibit a lower degree of oxidation than those collected
during Spring and Summer seasons. This is also in agreement with the assumption that
atmospheric WSOM has different origins depending of the season, with Summer values
probably having a higher contribution from photochemical oxidation processes (SOA
formation) and Winter values resulting from direct emissions of oxygenated particulate
organics from various anthropogenic sources, including biomass burning. According to
Krivacsy et al. (2001a), an O/C atomic ratio of 0.55 indicates that the organic structures
have a high degree of oxidation. In this study, the aerosol WSOC hydrophobic acids show
O/C atomic ratios lower than the value reported by Krivacsy et al. (2001a). Moreover, the
values obtained in the present study are lower than those reported by Duarte et al. (2007)
for a rural location near the city of Aveiro (0.41-0.55), thus suggesting that the urban
aerosol WSOC hydrophobic acids exhibit a lower degree of oxidation than the rural
aerosol WSOM.
Table V-1 also shows the atomic N/C ratios for the WSOC hydrophobic acids
fractions from the urban aerosol samples. The N/C ratios exhibit a seasonal variation, with
the higher values being observed for aerosol samples collected during Spring and Summer
A 2010 seasons. Duarte et al. (2007) also reported a similar annual trend for the atomic
N/C ratios of WSOC hydrophobic acids fractions from rural aerosol samples. The authors
suggested that the higher N/C ratio of samples collected during the warmer periods, i.e.
higher content in organic nitrogen compounds, could be associated with the enhancement
of photochemical reactions of individual vapor phase organic compounds with reactive
inorganic forms of nitrogen during these periods of high solar intensity. Surprisingly, the
N/C ratio for samples collected during the Summer B 2010 season is of the same order of
magnitude of those collected during Autumn and Winter seasons. The decrease in the value
of N/C for the Summer B 2010 samples could be the end result of the direct emission of
organic compounds from biomass burning processes during this period of intense forest
fires. This finding supports the assumption that in this particular warmer period, the
aerosol WSOC hydrophobic acids fractions have an important primary origin.
The elemental analysis data were also used for calculating a mass conversion factor
that will be used for converting the amount of aerosol WSOC into mass of aerosol WSOM.
Structural characterization of water-soluble organic matter from fine urban atmospheric aerosols
113
Using this approach an aerosol mass closure was achieved on the basis of carbonaceous
fractions alone (please see section 4.8 for additional details). The conversion factors
calculated for the aerosol WSOM samples collected at the city of Aveiro ranged between
1.5 and 1.8, with an average (and standard deviation) value of 1.6 (0.1). This parameter
was slightly higher for samples collected during the Spring and Summer seasons (1.7-1.8)
than for those collected during the colder seasons (1.5-1.6). The conversion factors
obtained in the present study are slightly lower than those reported by Duarte (2006) for
aerosol WSOM samples collected at a rural location (between 1.7 and 1.9, with an average
value of 1.8), somehow reflecting the influence of the different levels of pollution at both
sampling sites. Apparently, the conversion factor parameter varies over time and from site
to site; nevertheless, in the present study, a constant value of 1.6 was used to calculate the
mass of aerosol WSOM (additional information is provided in section 4.8).
5.4. FTIR-ATR spectroscopy of aerosol water-soluble organic carbon
hydrophobic acids
The FTIR-ATR spectra of the aerosol WSOC hydrophobic acids samples are shown
in Fig. V-5. Overall, the spectra exhibit predominantly the presence of oxygen containing
functional groups and aliphatic C-H groups, with the majority of the valence vibrations
being characteristics to all samples. These spectra are also very similar to those of aerosol
WSOC hydrophobic acids collected at a rural site near the city of Aveiro (Duarte et al.,
2007).
Chapter V
114
Fig. V-5. FTIR-ATR spectra (4000–600 cm-1
) of aerosol WSOC hydrophobic acids samples from the
different seasons.
Winter/Spring 2011
Winter 2011
Autumn B 2010
Autumn A 2010
Summer B 2010
Summer A 2010
Spring B 2010
Spring A 2010
Winter/Spring 2010
Autumn 2009
(cm-1
)
6009001200150018002100240027003000330036003900
Structural characterization of water-soluble organic matter from fine urban atmospheric aerosols
115
The interpretation of the FTIR-ATR spectra was based on the assignments given in
the literature for natural organic matter from different environmental matrices. As such, the
following infrared absorption bands are common to all spectra: 1) a broad band centred at
around 3340 cm-1
, which is usually assigned to O-H stretching of hydroxyl, carboxyl and
phenol groups (Bellamy, 1975; Santos and Duarte, 1998; Kiss et al., 2002; Duarte et al.,
2007): 2) the bands in the 2970-2840 cm-1
region, which are usually attributed to C-H
stretching of methyl (centred at around 2959 cm-1
) and methylene (centred at around 2893
cm-1
) groups of aliphatic chains (Santos and Duarte, 1998; Duarte et al., 2007); 3) the
strong band at near 1710 cm-1
(ranging from 1705 cm-1
to 1718 cm-1
), which is generally
assigned to unconjugated carbonyl (C=O) stretching mainly of carboxyl groups and, in a
lesser extent, of aldehydic and/or ketonic C=O groups (Santos and Duarte, 1998; Duarte et
al., 2007); 4) the band centred at near 1630 cm-1
, which is usually attributed to the CC
stretching of aromatic rings and to the C=O stretching of conjugated carbonyl groups
(Santos and Duarte, 1998). It is noteworthy that bending vibration of water is centered at
1640 cm-1
and may also contribute to infrared absorption in this region if the sample is not
thoroughly dried (Duarte et al., 2007); 5) the band centred at approximately 1200 cm-1
,
which is assigned to O-H bending and C-O stretching vibrations mainly of carboxyl groups
(Santos and Duarte, 1998), and finally, 6) the band at approximately 853 cm-1
, which is
usually assigned to out-of-plane deformation vibrations of C=C bonds of alkanes (Bellamy,
1975).
One of the major differences between the spectra of the aerosol WSOC
hydrophobic acids lays in the band at around 1630 cm-1
, whose relative intensity is higher
in the spectra of the Spring B 2010, Summer A 2010, and Autumn A 2010 samples. In the
spectrum of the Spring A 2010 sample, the intensity of such band appears to be of the same
order of magnitude of that centred at 1710 cm-1
. Another difference is the presence of a
weak absorption band centred at around 1512-1517 cm-1
in the FTIR-ATR spectra of the
WSOC hydrophobic acids from Autumn 2009, Winter/Spring 2010, Autumn B 2010,
Winter 2011, and Winter/Spring 2011 samples. A band at this same wavenumber is typical
and quite strong in spectra of lignins (Santos et al., 2000; Duarte et al., 2003) and is due to
the C-C stretching vibrations of aromatic rings of siringyl and guaiacyl units (Fengel and
Wegner, 1984; Faix, 1992). These spectra also show a very weak band at approximately
1333-1340 cm-1
, which is usually attributed to ring breaking vibrations of siringyl units in
Chapter V
116
lignin spectra (Fengel and Wegner, 1984). The presence of these aromatic signals due to
lignin derived structures corroborates those found in the CPMAS 13
C NMR spectroscopy
(section 5.5), highlighting the contribution of wood burning processes into the chemical
properties of the WSOC fractions during the colder periods. A similar FTIR-ATR
signature was also found in the FTIR spectra of WSOC hydrophobic acids collected at a
rural site during Autumn, Autumn/Winter and Winter seasons (Duarte et al., 2007).
Another feature observed in the FTIR-ATR of the WSOC fractions from the Spring A
2010, Spring B 2010, and Summer A 2010 seasons is a week band at near 1575 cm-1
. This
band is not shown in spectra of other WSOC samples and, to the best of our knowledge,
this band is unusual in infrared spectra of aerosol WSOM from other sites. According to
Coates (2006), N-H bending vibrations of secondary amines could contribute to absorption
at this frequency. An additional feature observed in the FTIR-ATR spectra of the WSOC
hydrophobic acids samples collected during the colder periods (Autumn, Winter, and
Winter/Spring) is the presence of a weak absorption band between1449 and 1456 cm-1
,
usually attributed to asymmetric bending vibrations of C–H bonds of methyl and
methylene groups of aliphatic chains (Bellamy, 1975).
5.5. CPMAS 13
C NMR spectroscopy of aerosol water-soluble organic
carbon hydrophobic acids
This section is devoted to the investigation of the main structural features of aerosol
WSOC hydrophobic acids samples collected during sampling campaign II in the city of
Aveiro by means of solid-state cross polarization with magic angle spinning 13
C NMR
(CPMAS 13
C NMR) spectroscopy. Due to the low amount of solid residue of WSOC
hydrophobic acids fractions at the end of the freeze-drying procedure, only five aerosol
WSOM samples were characterized by this technique, namely those collected in Autumn
2009, Summer B 2010, Autumn B 2010, Winter 2011, and Winter/Spring 2011 seasons.
Fig. V-6 shows the solid-state CPMAS 13
C NMR spectra of these five aerosol WSOC
hydrophobic acids samples.
Structural characterization of water-soluble organic matter from fine urban atmospheric aerosols
117
Fig. V-6. Solid-state CPMAS 13
C NMR spectra of WSOC hydrophobic acids fractions of five aerosol
samples representative of different seasonal periods.
In accordance with the assignments described in the literature for natural organic
matter from various sources, including aerosol WSOM from other locations (Malcolm,
Chemical shift (ppm)
-20020406080100120140160180200220240
Autumn 2009
Summer B 2010
Autumn B 2010
Winter 2011
Winter/Spring 2011
Chapter V
118
1989; Santos et al., 2000; Lambert and Lankes, 2002; Duarte et al., 2005, 2007, 2013), four
major regions were considered in these CPMAS 13
C NMR spectra: 0-50 ppm
(unsubstituted saturated aliphatic carbons), 60-95 ppm (aliphatic carbons singly bonded to
one oxygen or nitrogen atom), 110-160 ppm (aromatic and unsaturated carbons), and 160-
190 ppm (carboxyl, ester and amide carbons). Minor peaks are also seen at 50-60 ppm
(carbons of methyl groups of methyl ethers), 95-110 ppm (aliphatic carbons bonded to two
oxygen atoms, such as anomeric carbons of polysaccharides), and 190-230 ppm (carbonyl
carbons of aldehydes and ketones). Table V-2 gives the 13
C NMR estimates (as percentage
peak areas) of carbon distribution in WSOC hydrophobic acids.
Table V-2. Percentage distribution of carbon in aerosol WSOC hydrophobic acids based on solid-state
CPMAS 13
C-NMR analysis. “N.D.” refers to NMR signal not detected.
Sample
Percentage peak area in each spectral region on the ppm scale
190-230 160-190 110-160 95-110 60-95 50-60 0-50
Autumn 2009 5.3 9.4 25.3 2.7 10.1 5.4 41.8
Summer B 2010 5.4 12.0 11.4 3.2 14.9 N.D. 53.2
Autumn B 2010 6.8 10.0 24.5 3.5 11.0 5.2 38.9
Winter 2011 6.9 10.3 25.6 3.7 11.6 5.5 36.4
Winter/Spring
2011 6.5 10.9 20.7 3.5 13.2 4.2 41.0
The CPMAS 13
C NMR spectra in Fig. V-6 reveal that the various aerosol WSOC
hydrophobic acids samples hold similar carbon functional groups; however, they differ in
terms of the relative carbon distribution (as seen in Table V-2). The most noticeable
feature in these spectra is the strong and broad aliphatic carbon signal (36.4-53.2% of the
total NMR peak area). Furthermore, signal intensities for aliphatic carbons in WSOC
hydrophobic acids vary with the sampling period, being slightly lower in samples collected
in Autumn and Winter seasons compared to those collected during the Summer period. The
presence of three peaks centred at 21 ppm, 25 ppm, and 29 ppm, with the first peak
attributed to methylene groups in alkyl chains and the other two peaks to methine groups in
Structural characterization of water-soluble organic matter from fine urban atmospheric aerosols
119
alkyl chains (Malcolm, 1989), are also seen in the spectra of WSOC hydrophobic acids
collected in the colder seasons. The spectrum of the sample collected in Winter 2011 also
exhibit an unresolved peak at 35 ppm, usually assigned to methylene carbons of branched
alkyl chains (Malcolm, 1989). This resonance appears to be shifted towards lower
chemical shift region, i.e., 33 ppm, in the spectra of the samples collected in Autumn and
Winter/Spring seasons. A possible assignment for this resonance is the presence of
methylene carbons , , , and from terminal methyl group (Hayes et al., 1989). The
peaks centred at 21 ppm, 29 ppm, and 35 ppm were also identified in WSOC hydrophobic
acids collected by Duarte et al. (2007) in Autumn and Winter seasons at a rural site.
The CPMAS 13
C NMR spectra of aerosol WSOC hydrophobic acids samples from
the colder seasons also exhibit resonance at approximately 55 ppm and 147 ppm. These
two peaks are absent from the spectrum of the WSOC hydrophobic acids collected in the
Summer season. A spectroscopic signature similar to those of peaks at 55 ppm and 147
ppm is typical in spectra of lignin structural units (Haw et al., 1984; Hatfield et al., 1987),
and they are also present, but considerably more intense, in the spectra of humic and fulvic
acids from Kraft pulp mill effluents (Duarte et al., 2003). While peak at 55 ppm is due to
methoxyl groups such as those of syringyl and guaiacyl units, the signal at 147 ppm is
attributed to oxygen-substituted aromatic ring carbons (Haw et al., 1984; Hatfield et al.,
1987). Biomass combustion processes appears to be a definite and significant source for
the presence of lignin derived structures in aerosol WSOC hydrophobic acids samples of
colder seasons. During these periods the air temperature reached low values leading to an
increase in domestic heating, mainly by means of wood combustion processes in domestic
fireplaces. Besides levoglucosan and related anhydrosaccharides (e. g. mannosan and
galactosan), which are often used as tracers for smoke particulate matter from burning of
biomass, fine aerosol fractions of smoke samples can also contain thermally unaltered and
partially altered lignin-derived compounds, such as aromatic phenols (e.g. siringyl and
guaiacyl derivatives), aldehydes, ketones, acids and alcohols (Simoneit et al., 1993; Rogge
et al., 1998). Lignin pyrolysis products have generally the same substituent pattern (OH,
OCH3) on the aromatic rings as the precursor aromatic alcohols from which they were
derived (Simoneit et al., 1993). Wood burning can also explain the higher aromatic content
(total percentage peak areas in the 160-110 ppm range) of the WSOC hydrophobic acids
samples collected in colder seasons (20.7-25.6%) compared to that collected in the
Chapter V
120
Summer period (11.4%). These results are in good agreement with the studies of Duarte et
al. (2007, 2008b), where it has been clearly identified a biomass-burning fingerprint in
aerosol WSOC samples collected during Fall/Winter seasons at a rural location. In the
Spring/Summer seasons, on the other hand, the same authors suggested that the WSOC
might be due to secondary atmospheric oxidation processes, thus indicating a change in the
sources which are active in the various seasons. Samburova et al. (2005) likewise
suggested the primary emissions from wood combustion during Winter months, and
photochemical production during the Summer months, as possible sources for atmospheric
WSOC in an urban area. Sannigrahi et al. (2006) also reported a measurable amount of
aromatic carbon in WSOC hydrophobic acids fraction from urban atmospheric aerosols
collected during the Summer season. The authors suggested that the aromatic structures
present in their samples are likely to be mainly a product of motor vehicle emissions or
SOA-producing reactions. Taking into account these reports, one may also suggest that
motor vehicle exhaust and SOA formation could be possible sources of aromatic carbon to
the aerosol WSOC hydrophobic acids collected during Summer 2010 season at the city of
Aveiro.
All spectra also exhibit a week resonance at approximately 52 ppm. Using two-
dimensional liquid NMR spectroscopy for characterizing an aerosol WSOC sample
collected during the Winter season, Duarte et al. (2008b) found evidence of the presence of
a 13
C signal at 52 ppm, which the authors associated to amino sugar residue containing a
single ketoamide group within the hexose unit. Baldock et al. (1992) also reported that
amino carbon of protein structures resonate in the 45-65 ppm chemical shift region. The
resonance in the 60–95 ppm region is also common to all spectra (10.1-14.9% of the NMR
peak area) and it is likely to originate from the various HC–OH fragments of cellulose or
other carbohydrates structures (Hayes et al., 1989; Keeler et al., 2006; Duarte et al., 2008b,
2013; Simpson and Simpson, 2009). The presence of a downfield resonance at
approximately 104 ppm, which is representative of anomeric C, further corroborates the
presence of carbohydrate-like moieties in the aerosol WSOC hydrophobic fractions.
Unsubstituted aromatic C also show resonance within this spectral region (Duarte et al.,
2008), thus suggesting that 13
C NMR intensity in the 95–110 ppm region could also have a
small contribution from such structural moieties.
Structural characterization of water-soluble organic matter from fine urban atmospheric aerosols
121
Integration of the 160-190 ppm spectral region reveals that 9.4% to 12% of the
carbon in WSOC hydrophobic acids is associated with carboxylic acid functional groups
(Fig. V-6; Table V-2), with the highest value being obtained for samples collected during
the Summer 2010 season. As early mentioned, ester and amide carbons are also likely to
contribute to resonance in this spectral region. All spectra of the WSOC hydrophobic acids
also show evidences for the presence of carbonyl carbons of aldehydes and ketones (190-
230 ppm range), which accounts to 5.3-6.9% of the total NMR peak area.
5.6. Conclusions
This chapter presents the structural characterization of the aerosol WSOC
hydrophobic acids by means of UV-Vis, EEM fluorescence, FTIR-ATR, and CPMAS 13
C
NMR spectroscopies, and elemental analysis. In general, the results obtained by the
different techniques showed differences among the WSOM samples from the different
seasons.
The results obtained with UV-Vis spectroscopy, namely the E2/E3 ratio and ε280,
suggest the existence of a seasonal trend in the values of both parameters despite the
similarity of the UV-Vis spectra. The E2/E3 ratios also suggested that the WSOC
hydrophobic acids of the aerosol samples collected in the Summer A 2010 and Summer B
2010 seasons exhibit a lower molecular size and a lower degree of aromaticity than those
collected in Autumn 2009, Winter/Spring 2010, Autumn B 2010, Winter 2011, and
Winter/Spring 2011. The ε280 values also revealed that the WSOC samples collected during
Summer B 2010 period are likely to be more enriched in unsaturated bond systems, which
could be a consequence of the impact of local forest fires registered during this period.
The EEM spectra of the aerosol WSOC samples exhibit four distinct fluorophores,
being this fluorescence signature observed in all spectra of the samples collected in
sampling campaign II. Additionally, the synchronous fluorescence spectra with a λ of 60
nm collected in the colder periods (Autumn 2009, Autumn B 2010, and Winter 2011
seasons) exhibit fluorescence at λExc 280 nm, suggesting the presence of lignin-derived
structures resulting from wood burning processes in domestic fireplaces for house heating
Chapter V
122
during those periods. The synchronous fluorescence spectra of the aerosol WSOC samples
collected in Summer B 2010 period, also shows such fluorescence band which may be
associated with the occurrence of forest fires. These features support the assumption that
the direct emission into the atmosphere of organic compounds from biomass burning
processes plays an important role in the bulk chemical properties of WSOC fractions.
In regard to the elemental analysis, results suggest that the values of C, H, N, and O
for the WSOC hydrophobic acids fractions were in the range of 57.9-65.7%, 5.61-6.32 %,
2.34-4.52%, and 25.9-31.4%, respectively, with no distinct seasonal variation.
Nevertheless, WSOC hydrophobic acids from Spring and Summer apparently have a slight
decrease in the relative quantity of C with simultaneous increase in the relative quantities
of H, N, and O. Results also suggest that the aerosol WSOC hydrophobic acids samples
collected during the colder seasons exhibit a lower degree of oxidation than those collected
during Spring and Summer seasons.
In general, the FTIR-ATR spectra exhibit predominantly the presence of oxygen
containing functional groups and aliphatic C-H groups, with the majority of the valence
vibrations being characteristics to all samples. CPMAS 13
C NMR data showed that the
WSOC hydrophobic acids from atmospheric aerosols are mostly aliphatic in nature,
followed by aromatics, oxygenated alkyls, and carboxylic acid functional groups.
Moreover, the aerosol WSOC hydrophobic acids collected during colder periods have a
higher aromatic content than those collected during the Summer season. The presence of
signals typical of lignin-derived structures (e.g., syringyl and guaiacyl units) in both
CPMAS 13
C NMR and FTIR-ATR spectra of aerosol WSOC hydrophobic acids samples
collected during colder seasons highlights the major contribution of wood burning
processes during these periods of low temperatures conditions into the bulk chemical
properties of WSOM from atmospheric particles.
123
VI
Comprehensive two-dimensional liquid
chromatography of water-soluble organic matter
from fine urban atmospheric aerosols
Chapter VI
124
Comprehensive two-dimensional liquid chromatography of water-soluble organic matter
from fine urban atmospheric aerosols
125
6.1. Introduction
The comprehensive two-dimensional liquid chromatography (LC × LC) technique
combines two liquid chromatographic modes with different separation mechanisms, where
all components in a sample mixture are subjected to the two separation modes. The 1st
dimension and 2nd
dimension columns are connected in series by means of a transfer
system (i.e., a switching valve equipped with two identical loops) located between them.
The function of this interface is to cut and then re-inject continuous fractions of the 1st
dimension column effluent onto the fast separation 2nd
dimension column. In order to
achieve comprehensive analysis and to preserve the separation in the 1st dimension, the
fractions injected onto the 2nd
dimension column must undergo elution before the
following re-injection. The retention times in the 2nd
dimension separation mode must be
equal or less than the duration of the modulation period (i.e., time during which a fraction
of the 1st dimension column effluent is being accumulated in one of the loops of the
switching valve). The 2nd
dimension column is usually connected to a detection system (a
single detector or detectors connected in series, depending on the purpose of the analysis),
such as a DAD, fluorescence detector (FLD), evaporative light scattering detector (ELSD),
or MS detector.
An important practical principle of LC × LC concerns the orthogonality of the LC
modes, i.e. the selectivity of the two separation modes must be completely independent
form each other and separate two properties of a given sample in a single run without
influencing themselves. As recently reviewed by Duarte et al. (2012), several approaches
to comprehensive LC × LC have been described in the literature, including IEX × reversed
phase (RP), IC × RP, RP × SEC, PALC × SEC, or normal phase (NP) × RP. Although
orthogonal (independent), the combination of NP with RP is probably the most difficult to
accomplish due to the immiscibility of the mobile phases at the interface of both separation
modes, which can cause broad and distorted peaks (Stoll et al., 2007; Dugo et al., 2008).
According to Dugo et al. (2008), the great advantage of LC × LC relatively to the one-
dimensional LC (1D-LC) is the maximum number of compounds that can be separated in a
single chromatographic run. Besides an increased peak capacity, LC × LC also generates
increased selectivity and resolution. As recently reviewed by Matos et al. (2012), data
Chapter VI
126
processing of LC × LC is also a rapid evolving subject since there is a general lack of
commercial software associated to analytical instrumentation. According to Matos et al.
(2012), the first algorithms developed for data processing in LC × LC were generalizations
of concepts from 1D chromatography; nowadays, there are already methods for non-
targeted and targeted analyses fully developed to deal with multidimensional
chromatographic data (Matos et al., 2012).
LC × LC, coupled to different detectors, has proven to be useful for the separation
of a large number of components in several complex matrices, such as the separation and
identification of compounds in Chinese medicines (Chen et al., 2004), analyses of
triacylglycerols in natural lipidic matrixes (Dugo et al., 2006), and analyses of antioxidant
phenolic acids in herb extracts (Kivilompolo and Hyötyläinen, 2007) and in wines (Dugo
et al., 2009; Matos et al., 2013). For the analysis of complex environmental samples, only
three studies have been produced thus far concerning the application of LC × LC, namely
the analysis of acidic compounds in atmospheric aerosols (Pól et al., 2006) and the
separation of NOM from aquatic (Duarte et al., 2012) and atmospheric particles (Barros,
2011). Pól et al., (2006) used a LC x LC method, consisting of a strong cation-exclusion
column in the 1st dimension and a RP column in the 2
nd dimension, coupled to a TOF-MS
detection system for the quantitative determination of carboxylic acids in methanolic
extracts of rural and urban atmospheric aerosols. Duarte et al. (2012) used a mixed-mode
hydrophilic interaction liquid chromatography (HILIC) column operated under aqueous RP
conditions in the 1st dimension and a SEC column in the 2
nd dimension, for mapping the
hydrophobicity versus molecular weight (MW) distribution of two well-known complex
organic mixtures: Suwannee River Fulvic Acids and Pony Lake Fulvic Acids. The LC ×
LC fractions were screened on-line by three detectors connected in series: DAD, FLD, and
ELSD. According to Duarte et al. (2012), the packing material of the mixed-mode HILIC
column used in their work features an alkyl long chain with hydrophilic diol functional
groups at the end, thus allowing this packing material to be used in either HILIC mode
(with organic-rich mobile phase) or RP mode (with water-rich mobile phase). In order to
distinguish the features of this chromatographic mode from traditional HILIC and RP, a
new term was employed by the authors: per aqueous liquid chromatography (PALC). This
designation was proposed by Pereira et al. (2009) in their study on the reversing of HILIC
mechanism into PALC retention mechanism for the separation of highly polar ionizable
Comprehensive two-dimensional liquid chromatography of water-soluble organic
matter from fine urban atmospheric aerosols
127
solutes. Thus, using a PALC × SEC method, Duarte et al. (2012) concluded that the
combination of two independent separation mechanisms is promising in extend the range
of NOM separation. For the cases where NOM separation was accomplished, smaller MW
group fractions seem to be related to a more hydrophobic nature. Regardless of the
detection method, the authors also concluded that the complete range of MW distribution
provided by their LC x LC method was lower than those reported in the literature for
Suwannee River Fulvic Acids and Pony Lake Fulvic Acids. It should also be mentioned
that Góra et al. (2012) have introduced an off-line method combining a RP column with a
SEC column for the fractionation of soil humic acids. This study totally differ from that of
Duarte et al. (2012) mostly because the second employed an on-line method, thus meaning
that the whole sample is subjected to separation in both chromatographic modes.
Conversely, Góra et al. (2012) collected only a small volume around the peaks maximum
obtained in the RP separation procedure before off-line re-injection into the SEC column.
It must also be emphasized that Góra et al. (2012) did not provided any information about
the size distribution of the soil humic acids, neither demonstrated how the different size
fractions relate with their hydrophobic nature.
As already mentioned in section 2.5 (Chapter 2 of this thesis), Barros (2011) also
employed the PALC × SEC method for fractionating WSOM from atmospheric particles
collected during Winter season at a rural location (Duarte et al., 2007). Besides resolving
the chemical heterogeneity of the aerosol WSOM sample, the PALC × SEC method also
allowed the estimative of the average MW distribution of the sample (157-891 Da).
Overall, the results of Barros (2011) highlighted the huge potential of LC x LC as a
promising tool for resolving the chemical heterogeneity of the complex WSOM in fine
atmospheric aerosols. Therefore, in this chapter, the PALC x SEC technique developed by
Barros (2011) and Duarte et al. (2012) was employed for: (i) resolving the chemical
heterogeneity of the urban aerosol WSOC hydrophobic acids samples characterized in
Chapter 5; (ii) determining how size-distinguished fractions differ in hydrophobicity; and
(iii) assessing the MW properties of the studied WSOM samples in regard to number (Mn)
and weight (Mw) average MW, and polydispersity (Mw/Mn).
Chapter VI
128
6.2. Development of PALC × SEC method for the analysis of urban
aerosol WSOC hydrophobic acids
As described by Duarte et al. (2012), when performing LC x LC analysis of
complex mixtures, the first procedure must entail the setup of the experimental conditions
of both chromatographic dimensions in regard to mobile phase composition, flow rate,
modulation period, and total analysis time. Besides mobile phase compatibility, important
for avoiding adsorption phenomena of the analytes into the 2nd
dimension column under
the injection plug (i.e. transfer volume) conditions, the speed of analysis in the 2nd
dimension will determine the sampling rate of the effluent from the 1st dimension column
(i.e. the modulation period) and, ultimately, the time needed for accomplishing a high
resolved LC x LC map of the sample (Duarte et al., 2012).
Although the PALC× SEC method employed in this study was based on the works
of Barros (2011) and Duarte et al. (2012), the separation of the urban aerosol WSOC
hydrophobic acids samples was first screened in the 1st and 2
nd dimension columns,
independently. This important step was necessary to set the flow rate at the 1st dimension,
the modulation period, the number of fractions of the effluent from the 1st dimension
column that were transferred into the 2nd
dimension column, and the total time of analysis,
thus aiming at ensure a good performance of the PALC x SEC method for resolving the
chemical heterogeneity of the urban aerosol WSOM samples.
Fig. VI-1 and Fig. VI-2 show the 1D chromatograms obtained under PALC and
SEC conditions, respectively, for the urban aerosol WSOC hydrophobic acids samples
collected in Autumn 2009, Summer B 2010, Autumn B 2010, Winter 2011, and
Winter/Spring 2011 seasons. For method development purposes in the 1st dimension, and
to avoid long elution times (almost three hours) in this first procedure, the flow rate used in
PALC separation was set at 0.5 mL min-1
, which is 25 times higher than the flow rate
actually applied in the PALC × SEC separations. It was assumed that the flow rate had
negligible effects on the chromatographic profile of the WSOC hydrophobic acids samples
in the 1st dimension. Three and two replicates of PALC and SEC separation of each sample
were performed, respectively; however, for better visualization, only the chromatograms
corresponding to the first replica obtained in each separation technique are shown.
Comprehensive two-dimensional liquid chromatography of water-soluble organic
matter from fine urban atmospheric aerosols
129
Fig. VI-1. One-dimensional (1D) chromatograms of the urban aerosol WSOC hydrophobic acids obtained
with the PALC technique, and recorded by different detection methods: (A) DAD operating at
254 nm, FLD at (B) λExc/λEm = 240/410 nm and (C) λExc/λEm = 320/415 nm, and (D) ELSD.
Additional details about the chromatographic conditions in terms of mobile phase composition
can be found in section 3.10 (Chapter 3). The flow rate in PALC was 0.5 ml min−1
.
Norm
alize
d UV
254n
m si
gnal
(a.u
)
0.0
0.2
0.4
0.6
0.8
1.0
Winter/Spring 2011
Winter 2011
Autumn B 2010
Summer B 2010
Autumn 2009
Norm
alize
d FL
240
/410
nm
sign
al (a
.u)
0.0
0.2
0.4
0.6
0.8
1.0
Winter/Spring 2011
Winter 2011
Autumn B 2010
Summer B 2010
Autumn 2009
Retention time (min)
0 5 10 15 20
Norm
alize
d EL
SD si
gnal
(a.u
)
0.0
0.2
0.4
0.6
0.8
1.0
Winter/Spring 2011
Winter 2011
Autumn B 2010
Summer B 2010
Autumn 2009
Norm
alize
d FL
320
/415
nm
sign
al (a
.u)
0.0
0.2
0.4
0.6
0.8
1.0
Winter/Spring 2011
Winter 2011
Autumn B 2010
Summer B 2010
Autumn 2009
A
B
C
D
Chapter VI
130
Fig. VI-2. One-dimensional (1D) chromatograms of the urban aerosol WSOC hydrophobic acids obtained
with the SEC technique, and recorded by different detection methods: (A) DAD operating at 254
nm, FLD at (B) λExc/λEm = 240/410 nm and (C) λExc/λEm = 320/415 nm, and (D) ELSD.
Additional details about the chromatographic conditions in terms of mobile phase composition
can be found in section 3.10 (Chapter 3). The flow rate in SEC was 2.5 ml min−1
.
Nor
mal
ized
UV
254n
m si
gnal
(a.u
)
0.0
0.2
0.4
0.6
0.8
1.0
Winter/Spring 2011
Winter 2011
Autumn B 2010
Summer B 2010
Autumn 2009
Nor
mal
ized
FL
240/
410
nm si
gnal
(a.u
)
0.0
0.2
0.4
0.6
0.8
1.0
Winter/Spring 2011
Winter 2011
Autumn B 2010
Summer B 2010
Autumn 2009
Retention time (min)
0 1 2 3 4 5
Nor
mal
ized
ELS
D si
gnal
(a.u
)
0.0
0.2
0.4
0.6
0.8
1.0
Winter/Spring 2011
Winter 2011
Autumn B 2010
Summer B 2010
Autumn 2009
Nor
mal
ized
FL
(320
/415
nm
) sig
nal (
a.u)
0.0
0.2
0.4
0.6
0.8
1.0
Winter/Spring 2011
Winter 2011
Autumn B 2010
Summer B 2010
Autumn 2009
A
B
C
D
Comprehensive two-dimensional liquid chromatography of water-soluble organic
matter from fine urban atmospheric aerosols
131
Overall, a fractionation of the urban aerosol WSOC hydrophobic acids has occurred
on the 1st dimension mixed-mode HILIC column, operated under PALC conditions, as
suggested by the presence of unresolved peaks between approximately 3.5 and 7.0 min in
the chromatographic profiles acquired by both UV and fluorescence detection. In terms of
ELSD detection, the obtained chromatograms are more featureless when compared to those
of the other two detectors. Unlike UV and molecular fluorescence, ELSD detection is
independent of the optical properties (i.e. type of chromophores) of the organic
components contained in the WSOM samples, being only dependent on their concentration
(Duarte et al., 2012). Therefore, the results obtained under PALC conditions reported in
Fig. VI-1 suggest that the relative contribution to absorbance and quantum yield of
fluorescence of the organic compounds with highly conjugated bond systems is higher
than that of ELSD, thus giving to some extent more feature-rich chromatograms.
In what concerns the SEC chromatograms shown in Fig. VI-2, most of the aerosol
WSOC hydrophobic acids samples elute from the SEC column in about 150 s (2.5 min),
which was set as the modulation period. Nevertheless, the fluorescence chromatograms
obtained at λExc/λEm of 240/410 nm and 320/415 nm, also show that some minor
fluorophores are still eluting from the SEC column at retention times higher than 2.5 min.
Despite this observation, it was decided to establish the retention time of 2.5 min as the
modulation period for two main reasons: 1) the SEC profiles of Fig. VI-2 were obtained for
the whole sample, whereas in the LC× LC method, we are dealing with fractions from the
1st dimension column effluent, which experience some mixing and dilution with the mobile
phase of the 2nd
dimension, thus arriving with a somewhat lower concentration at the SEC
column and subsequent detection systems; and 2) setting the modulation period for values
higher than 4 min (suggested by fluorescence detection at λExc/λEm = 240/410 nm) will
yield a total time of analysis in the LC× LC method of 600 min (10 hours), which is totally
unfeasible in practical terms. With a modulation period of 150 s, the total time of analysis
in the LC× LC method was set as 375 min. As depicted in Fig. VI-2, the SEC profiles of
the aerosol WSOC hydrophobic acids show a unimodal distribution under the selected
chromatographic conditions, although the ELSD chromatogram of the sample collected
during the Winter/Spring 2011 season exhibit an additional unresolved peak at 1.5 min.
Furthermore, the UV, fluorescence at λExc/λEm = 240/410 nm, and ESLD chromatograms of
Chapter VI
132
this sample are slightly shifted towards higher retention times as compared to those of the
other samples. This feature suggests that the WSOC hydrophobic acids from the
Winter/Spring 2011 season may encompass organic structures with somewhat lower
molecular weight than those of the other aerosol WSOM samples.
6.3. Analysis of urban aerosol WSOC hydrophobic acids through
PALC × SEC methodology
The contour plots obtained for the WSOC hydrophobic acids separations using the
PALC × SEC method are shown in Fig. VI-3 to Fig. VI-7. The fractions identified in each
contour plot (labeled by numbers), their respective retention times in the 1st dimension, and
the MW characteristics (Mn, Mw and Mw/Mn) generated from the PALC x SEC analyses
of the WSOM samples are listed in Table VI-1 to Table VI-3. In this work, the Mn and
Mw values were estimated using the PSS WinGPC Unity software package, after the
calibration of the SEC column with sodium polystyrenesulfonate standards and HPLC
grade acetone 5% (v/v) (section 3.10, Chapter 3).
As it can be seen in Fig. VI-3 to Fig. VI-7, the PALC × SEC method allowed the
separation of the WSOC hydrophobic acids samples into several fractions, where the
contour plots of UV absorption, fluorescence, and ELSD data display a different elution
pattern not only for the same sample but also between the different samples. For example,
the UV absorption contour plots of WSOM samples collected in Summer B 2010,
Winter/Spring 2011, and Autumn 2009, exhibit 4, 6, 7 chromatographic peaks,
respectively, while samples collected in Autumn B 2010 and Winter 2011 both have 8
chromatographic peaks. The only exception was verified for the ELSD data, where
apparently no fractionation has been achieved and just a single broad chromatographic
peak can be observed. As already stated (section 6.2), the ELSD detection is independent
of the optical properties of the organic components contained in the WSOM samples. In
this sense, the obtained ELSD data are less prone to be influenced by the unequal molar
absorptivities and fluorescence properties of the organic moieties.
Comprehensive two-dimensional liquid chromatography of water-soluble organic matter from fine urban atmospheric aerosols
133
Fig. VI-3. PALC × SEC contour plots of aerosol WSOC hydrophobic acids from Autumn 2009 recorded by different detection methods (UV absorption at 254 nm,
fluorescence (FL) at λExc/λEm of 240/410 nm and 320/415 nm, and evaporative light-scattering (ELSD)). Colours represent the intensity of the analytical
signal.
UV – 254 nm
5
6
7
4
3
2
1
100
80
60
40
20
0
FL – 240/410 nm100
80
60
40
20
0
5
6
7
43
2
1
FL – 320/415 nm100
80
60
40
20
0
8
9
10
76
5
1
2
43
ELSD100
80
60
40
20
0
1
Chapter VI
134
Fig. VI-4. PALC × SEC contour plots of aerosol WSOC hydrophobic acids from Summer B 2010 recorded by different detection methods (UV absorption at 254 nm,
fluorescence (FL) at λExc/λEm of 240/410 nm and 320/415 nm, and evaporative light-scattering (ELSD)). Colours represent the intensity of the analytical
signal.
UV- 254 nm
2
3
4
1
100
80
60
40
20
0
FL - 240/410 nm100
80
60
40
20
0
4
5
3
2
1
FL - 240/410 nm
FL - 320/415 nm100
80
60
40
20
0
1
6
7
5
4 32
ELSD100
80
60
40
20
0
1
Comprehensive two-dimensional liquid chromatography of water-soluble organic matter from fine urban atmospheric aerosols
135
Fig. VI-5. PALC × SEC contour plots of aerosol WSOC hydrophobic acids from Autumn B 2010 recorded by different detection methods (UV absorption at 254 nm,
fluorescence (FL) at λExc/λEm of 240/410 nm and 320/415 nm, and evaporative light-scattering (ELSD)). Colours represent the intensity of the analytical
signal.
UV- 254 nm100
80
60
40
20
0
6
7
8
5
4
2
1
3
FL - 240/410 nm100
80
60
40
20
0
5
6
7
43
2
1
FL - 320/415 nm100
80
60
40
20
0
4
5
6
32
1
ELSD100
80
60
40
20
0
1
Chapter VI
136
Fig. VI-6. PALC × SEC contour plots of aerosol WSOC hydrophobic acids from Winter 2011 recorded by different detection methods (UV absorption at 254 nm,
fluorescence (FL) at λExc/λEm of 240/410 nm and 320/415 nm, and evaporative light-scattering (ELSD)). Colours represent the intensity of the analytical
signal.
UV- 254 nm
6
7
8
4
5
2
1
3
100
80
60
40
20
0
FL - 240/410 nm100
80
60
40
20
0
7
6
5
3
4
2
1
100
80
60
40
20
0
FL - 320/415 nm
4
5
3
2
1
ELSD 100
80
60
40
20
0
1
Comprehensive two-dimensional liquid chromatography of water-soluble organic matter from fine urban atmospheric aerosols
137
Fig. VI-7. PALC × SEC contour plots of aerosol WSOC hydrophobic acids from Winter/Spring 2011 recorded by different detection methods (UV absorption at 254
nm, fluorescence (FL) at λExc/λEm of 240/410 nm and 320/415 nm, and evaporative light-scattering (ELSD)). Colours represent the intensity of the
analytical signal.
UV- 254 nm100
80
60
40
20
0
6
4
5
2
3
1
FL - 240/410 nm100
80
60
40
20
0
5
2
3
4
1
FL - 320/415 nm100
80
60
40
20
0
6
2
3
4
1
5
ELSD100
80
60
40
20
0
1
Chapter VI
138
Table VI-1. Molecular weight characteristics of the aerosol WSOC hydrophobic acids from Autumn 2009 and Autumn B 2010, using the PALC × SEC methodology.
Detector Peak
number
Autumn 2009 Autumn B 2010
Elution Time
1st Dimension (min)
Mn (Da) Mw (Da) Mw/Mn Elution Time
1st Dimension (min)
Mn (Da) Mw (Da) Mw/Mn
UV - 254 nm
1 38 525 614 1.17 37 687 711 1.04
2 47 422 539 1.28 47 477 570 1.19
3 71 323 391 1.21 52 457 505 1.10
4 82 261 309 1.18 70 480 516 1.08
5 106 89 94 1.06 82 438 467 1.07
6 116 85 91 1.07 105 78 79 1.01
7 123 75 78 1.04 116 78 79 1.01
8
121 74 75 1.01
FL - 240/410 nm
1 43 614 710 1.16 43 689 827 1.20
2 50 376 565 1.50 50 347 549 1.58
3 64 438 547 1.25 65 472 529 1.12
4 71 425 548 1.29 70 368 506 1.37
5 80 308 394 1.28 80 304 354 1.16
6 84 264 384 1.45 84 251 343 1.37
7 148 45 46 1.01 97 333 357 1.07
FL - 320/415 nm
1 35 702 854 1.22 35 953 1100 1.15
2 40 543 604 1.11 37 1137 1241 1.09
3 46 557 641 1.15 46 978 1124 1.15
4 50 250 405 1.62 50 330 496 1.50
5 57 563 597 1.06 64 532 565 1.06
6 64 369 474 1.29 70 1111 1125 1.01
7 70 397 523 1.32
8 77 531 558 1.05
9 85 314 361 1.15
10 149 45 45 1.00
ELSD 1 39 158 177 1.12 40 146 158 1.09
Comprehensive two-dimensional liquid chromatography of water-soluble organic matter from fine urban atmospheric aerosols
139
Table VI-2. Molecular weight characteristics of the aerosol WSOC hydrophobic acids from Summer B 2010, using the PALC × SEC methodology.
Detector Peak number
Summer B 2010
Elution Time
1st Dimension (min)
Mn (Da) Mw (Da) Mw/Mn
UV - 254 nm
1 40 253 500 1.98
2 47 774 942 1.22
3 68 661 929 1.41
4 75 597 702 1.18
FL - 240/410 nm
1 89 464 547 1.18
2 99 437 738 1.69
3 121 483 557 1.16
4 129 449 492 1.09
5 135 256 393 1.54
FL - 320/415 nm
1 37 866 1064 1.23
2 45 818 986 1.20
3 47 799 1021 1.28
4 52 387 610 1.58
5 70 505 885 1.75
6 75 491 619 1.26
7 84 391 509 1.30
ELSD 1 41 232 283 1.22
Chapter VI
140
Table VI-3. Molecular weight characteristics of the aerosol WSOC hydrophobic acids from Winter 2011 and Winter/Spring 2011, using the PALC × SEC methodology.
Detector Peak
number
Winter 2011 Winter/Spring 2011
Elution Time
1st Dimension (min)
Mn (Da) Mw (Da) Mw/Mn Elution Time
1st Dimension (min)
Mn (Da) Mw (Da) Mw/Mn
UV - 254 nm
1 43 832 889 1.07 37 257 296 1.15
2 47 553 692 1.25 52 151 222 1.48
3 54 535 622 1.16 71 188 212 1.13
4 70 458 580 1.27 80 153 169 1.11
5 76 370 437 1.18 90 109 140 1.29
6 105 106 110 1.03 174 48 48 1.01
7 115 102 106 1.04
8 119 94 96 1.02
FL - 240/410 nm
1 35 886 1047 1.18 37 428 524 1.22
2 39 743 895 1.20 52 277 452 1.63
3 50 564 867 1.54 71 290 358 1.23
4 65 654 815 1.25 80 282 317 1.12
5 70 584 771 1.32 90 182 258 1.41
6 80 535 587 1.10
7 84 361 517 1.43
FL - 320/415 nm
1 47 652 772 1.18 36 625 708 1.13
2 50 337 517 1.54 51 272 410 1.51
3 54 591 634 1.07 67 385 463 1.20
4 64 520 619 1.19 70 383 455 1.19
5 67 496 585 1.18 75 343 400 1.17
6 86 269 322 1.20
ELSD 1 40 209 220 1.05 40 127 138 1.09
Comprehensive two-dimensional liquid chromatography of water-soluble organic
matter from fine urban atmospheric aerosols
141
Regarding the MW distribution of the samples, and in terms of UV absorption
detection, the WSOC hydrophobic acids from Summer B 2010 have apparently higher Mw
than the other WSOM samples, ranging from 500 to 942 Da. The Autumn 2009 and
Autumn B 2010 show similar MW distribution, with the Mw values ranging approximately
from 75 to 711 Da, whereas the Mw values of the Winter 2011 and Winter/Spring 2011
samples ranged between 96 and 889 Da, and 48 and 296 Da, respectively. In the UV
contour plot of the Autumn 2009, Autumn B 2010, and Winter 2011 samples, there are
cluster of fractions that have low Mw values (78-110 Da), but different hydrophobicity
(indicated by the retention time in 1st dimension PALC separation). According to the work
of Duarte and Duarte (2011b), the occurrence of organic compounds with such a low Mw
range in these atmospheric aerosol samples may be indicative of the presence of freshly
emitted SVOCs with low Mw. Nonetheless, the presence of fractions exhibiting somewhat
higher Mw values (ranging from 300 to 890 Da) in these same samples could also be
associated with the emission to the atmosphere of organic particles from other sources than
those of SVOCs (e.g. wood burning). On the other hand, the presence of two fractions with
apparently high Mw in the Summer B 2010 sample (namely peaks 2 and 3, with Mw
values of 942 and 929 Da, respectively) may be explained by the long residence time in the
atmosphere of low Mw organic compounds that undergo oligomerization and/or
polymerization reactions mediated by solar radiation, thereby resulting in molecular
structures with higher Mw (Gelencser et al., 2002).
The contour plots of the WSOC hydrophobic acids samples obtained with
fluorescence detection exhibit between 5 and 10 chromatographic peaks (Fig. VI-3 to Fig.
VI-7, and Table VI-1 to Table VI-3). Apparently, all WSOC hydrophobic acids samples
contain a group of less hydrophobic fluorescent components (retention time in the 1st
dimension between 35 and 99 min at both λExc/λEm pairs) whose average Mw values range
between 738 and 1047 Da. The Autumn 2009 sample shows one additional more
hydrophobic fluorescent fraction at the retention time of 148-149 min in the 1st dimension,
whose average Mw is of approximately 45-46 Da. Using fluorescence detection at λExc/λEm
= 240/410 nm, it is also possible to observe three additional more hydrophobic fluorescent
fractions for the Summer B 2010 sample at the retention times in the 1st dimension
between 121 and 135 min, with average Mw values ranging from 393 to 557 Da. The
Chapter VI
142
ELSD detection method, on the other hand, generated only one chromatographic peak at
the retention time of approximately 40 min in the 1st dimension and with average Mw
values ranging from 138 to 283 Da. Findings suggest that the less hydrophobic acid
components appear to have larger Mw than the highly hydrophobic acid components.
Duarte et al (2012) also concluded that the fractions with low Mw have a more
hydrophobic nature; however, depending on the detector, this trend was not observed in all
samples analysed in this study. Results obtained in this study demonstrate that the PALC x
SEC method offers a new perspective on the structural heterogeneity of WSOM, allowing
to visualize differences between chromatographic fractions in terms of MW distribution
and hydrophobicity.
As it can be seen in Tables VI-1 to VI-3, and regardless of the detection system, the
Summer B 2010 sample has the highest values of the polydispersity index (Mw/Mn),
reaching a value of 1.98 for the fraction labeled as 1 in the UV contour plot. These results
suggest that the MW distribution of this Summer sample has higher heterogeneity than
those of the other samples. On the other hand, polydispersity values close to 1.0 highlight
the low heterogeneity of the MW distributions, thus suggesting that the PALC × SEC
method was successful in resolving the chemical heterogeneity of the aerosol WSOM
samples.
When comparing with the results reported in the literature, and only focusing on the
results obtained by UV absorption detection, the results obtained in this study for the urban
aerosol WSOC hydrophobic acids collected in Winter (Mw = 96-889 Da) are within the
range of those reported by Barros (2011) for rural aerosol WSOC hydrophobic acids (Mw
= 157-649 Da). Overall, the MW distribution of aerosol WSOM samples obtained in this
study by means of PALC × SEC method is in agreement with a previous study conducted
by Kiss et al. (2003), who reported Mw values of 200-300 Da for their aerosol WSOM
samples. In the same way, the obtained MW distribution is within the range of those
reported by Samburova et al. (2005), which used SEC/LDI-MS to confirm that the upper
mass limit of their aerosol WSOM sample reached 700 Da.
Comprehensive two-dimensional liquid chromatography of water-soluble organic
matter from fine urban atmospheric aerosols
143
6.4. Conclusions
An LC x LC method, combining a mixed-mode HILIC column operating under
PALC conditions in the 1st dimension and a SEC column in the 2
nd dimension, was applied
with purpose of resolving the chemical heterogeneity of WSOC hydrophobic acids from
atmospheric aerosols collected in different seasons. The following conclusions can be
drawn:
a) the MW distribution of the all WSOC hydrophobic acids ranged from 48 to 942
Da and from 45 to 1241 Da in terms of UV absorption and fluorescence
detection, respectively. The ELSD detection generated a MW distribution
ranging between 128 and 283 Da;
b) fractions with high Mw values encompass organic structures with less
hydrophobicity;
c) the occurrence of organic compounds with low Mw range (78-110 Da) in the
aerosol WSOC hydrophobic acids collected in Autumn 2009, Autumn B 2010,
and Winter 2011 periods may be indicative of the presence of freshly emitted
SVOCs with low Mw;
d) the PALC × SEC method was successful in resolving the chemical
heterogeneity of the aerosol WSOM samples. Both UV absorption and
fluorescence detection seem to be appropriate as detection systems for these
complex organic mixtures, whereas ELSD detection provides featureless
chromatographic information.
Chapter VI
144
145
VII
General conclusions
Chapter VII
146
General conclusions
147
This thesis aimed at investigating the structural composition of the WSOM from
fine aerosol particles typical of an urban location. In a first stage, the adsorption of SVOCs
onto quartz fibre filters during the collection of aerosol particles was assessed. A global
carbon balance was also performed, as well as a mass closure study for the whole mass of
fine aerosol collected at the city of Aveiro. The UV-Vis and EEM fluorescence spectra of
each WSOC sample were recorded, and the WSOC samples were assembled in different
groups according to similar meteorological parameters before the isolation/fractionation of
the assembled samples into WSOC hydrophobic and hydrophilic acids fractions.
Afterwards, the WSOC hydrophobic acids fractions were characterized by means of
elemental analysis, and FTIR-ATR and solid-state CPMAS 13
C NMR spectroscopies.
Additionally, the WSOC hydrophobic acids fractions were further analyzed by a PALC x
SEC methodology in order to resolve their chemical heterogeneity and to determine how
size-distinguished fractions differ in terms of their hydrophobicity. The obtained results
allowed the following conclusions:
a) Apparently, volatilisation/condensation processes of SVOCs may occur on the
quartz fibre filters, or on particles surface, during aerosol sampling;
b) In the city of Aveiro, inhalable particulate matter is largely composed of fine
particles, globally above 50 % by weight, except for Summer season. During
the cold seasons, the atmospheric concentration of suspended particles exhibit a
significant increase, specially the fine size range, which could be a consequence
of local anthropogenic activities (e.g., wood burning). For fine particles, the
carbonaceous fraction represents approximately 16 and 47 % of the aerosol
mass in the warm (Summer A 2010) and cold (Autumn 2009 and Winter 2011)
seasons, respectively, and less than 6 % of the carbonaceous matter consist of
EC;
c) The elemental analysis of WSOC hydrophobic acids showed that the aerosol
WSOC hydrophobic acids samples collected during the colder seasons exhibit a
lower degree of oxidation than those collected during Spring and Summer
seasons;
Chapter VII
148
d) The analysis of the aerosol WSOC hydrophobic acids by CPMAS 13
C NMR
suggests a more aromatic content of the samples collected during low-
temperature conditions. The presence of signals typical of lignin-derived
structures in the CPMAS 13
C NMR and FTIR-ATR spectra of WSOC
hydrophobic acids samples from Autumn 2009, Autumn B 2010, Winter 2011,
and Winter/Spring 2011 seasons highlights the importance of wood burning
processes for domestic heating in the bulk chemical properties of WSOM from
atmospheric aerosols.
e) Using a PALC x SEC method, it was found that the Mw distribution of aerosol
WSOM ranged from 48 to 942 Da and from 45 to 1241 Da in terms of UV
absorption and fluorescence detection, respectively. The WSOC hydrophobic
acids fractions from Autumn 2009, Autumn B 2010, and Winter 2011 samples
showed the presence of very low Mw components, which could be a
consequence of fresh emitted organic particles. It was also observed that the
components with smaller average Mw tend to have a more hydrophobic
character. The PALC x SEC method coupled to multiple detectors (DAD, FLD,
and ELSD) was successful in resolving the chemical heterogeneity of the
aerosol WSOM samples;
f) Finally, the results obtained in this study are of great importance for a better
characterization and understanding of the chemical composition of the WSOM
from fine urban atmospheric aerosols.
For future work, it is suggested: i) the application of sophisticated high-resolution
analytical techniques (namely, 2D-NMR spectroscopy and MALDI-TOF spectrometry) for
a more in-depth insight into the molecular composition of the whole WSOM from fine
urban aerosols; ii) the assessment of how the structural information of the aerosol WSOM
vary with hydrophobicity and MW profile by the synergistic application of LC x LC and
high-resolution spectroscopic and spectrometric techniques; and iii) unfolding the
molecular complexity of aerosol WSOM from other urban environments.
149
References
References
150
References
151
Abbt-Braun, G., Lankes, U., Frimmel, F.H., 2004. Structural characterization of aquatic
humic substances - The need for a multiple method approach. Aquat. Sci. 66, 151–
170.
Alves, C., Carvalho, A., Pio, C., 2002. Mass balance of organic carbon fractions in
atmospheric aerosols. J. Geophys. Res. Atmos. 107, 8345.
Alves, C., Pio, C., Duarte, A., 2001. Composition of extractable organic matter of air
particles from rural and urban Portuguese areas. Atmos. Environ. 35, 5485–5496.
Andracchio, A., Cavicchi, C., Tonelli, D., Zappoli, S., 2002. A new approach for the
fractionation of water-soluble organic carbon in atmospheric aerosols and cloud
drops. Atmos. Environ. 36, 5097–5107.
Asa-Awuku, A., Sullivan, A.P., Hennigan, C.J., Weber, R.J., Nenes, A., 2008.
Investigation of molar volume and surfactant characteristics of water-soluble organic
compounds in biomass burning aerosol. Atmos. Chem. Phys. 8, 799–812.
Baduel, C., Voisin, D., Jaffrezo, J.L., 2010. Seasonal variations of concentrations and
optical properties of water soluble HULIS collected in urban environments. Atmos.
Chem. Phys. 10, 4085–4095.
Baldock, J.A., Oades, J.M., Waters, A.G., Peng, X., Vassallo, A.M., Wilson, M.A., 1992.
Aspects of the chemical structure of soil organic materials as revealed by solid-
state13C NMR spectroscopy. Biogeochemistry 16, 1–42.
Barros, A.C.C.A. de, 2011. Natural organic matter: study by two-dimensional liquid
chromatography. Universidade de Aveiro.
Bellamy, L.I., 1975. The Infrared Spectra of Complex Molecules. Chapman and Hall,
London.
Birch, M.E., Cary, R.A., 1996. Elemental carbon-based method for monitoring
occupational exposures to particulate diesel exhaust. Aerosol Sci. Technol. 25, 221–
241.
References
152
Carvalho, A., Pio, C., Santos, C., 2003. Water-soluble hydroxylated organic compounds in
German and Finnish aerosols. Atmos. Environ. 37, 1775–1783.
Castro, L.M., Pio, C.A., Harrison, R.M., Smith, D.J.T., 1999. Carbonaceous aerosol in
urban and rural European atmospheres: estimation of secondary organic carbon
concentrations. Atmos. Environ. 33, 2771–2781.
Cavalli, F., Facchini, M.C., Decesari, S., Emblico, L., Mircea, M., Jensen, N.R., Fuzzi, S.,
2006. Size-segregated aerosol chemical composition at a boreal site in southern
Finland, during the QUEST project. Atmos. Chem. Phys. 6, 993–1002.
Chan, M.N., Choi, M.Y., Ng, N.L., Chan, C.K., 2005. Hygroscopicity of water-soluble
organic compounds in atmospheric aerosols: Amino acids and biomass burning
derived organic species. Environ. Sci. Technol. 39, 1555–1562.
Charlson, R.J., Schwartz, S.E., Hales, J.M., Cess, R.D., Coakley, J.A., Hansen, J.E.,
Hofmann, D.J., 1992. Climate Forcing by Anthropogenic Aerosols. Science (80-. ).
255, 423–430.
Chen, J., Gu, B.H., LeBoeuf, E.J., Pan, H.J., Dai, S., 2002. Spectroscopic characterization
of the structural and functional properties of natural organic matter fractions.
Chemosphere 48, 59–68.
Chen, J., Ying, Q., Kleeman, M.J., 2010. Source apportionment of wintertime secondary
organic aerosol during the California regional PM10/PM2.5 air quality study. Atmos.
Environ. 44, 1331–1340.
Chen, X., Kong, L., Su, X., Fu, H., Ni, J., Zhao, R., Zou, H., 2004. Separation and
identification of compounds in Rhizoma chuanxiong by comprehensive two-
dimensional liquid chromatography coupled to mass spectrometry. J. Chromatogr. A
1040, 169–178.
Chow, J.C., Watson, J.G., Lu, Z., Lowenthal, D.H., Frazier, C.A., Solomon, P.A.,
Thuillier, R.H., Magliano, K., 1996. Descriptive analysis of PM25 and PM10 at
regionally representative locations during SJVAQS/AUSPEX Atmos. Environ. .
References
153
Claeys, M., Graham, B., Vas, G., Wang, W., Vermeylen, R., Pashynska, V., Cafmeyer, J.,
Guyon, P., Andreae, M.O., Artaxo, P., Maenhaut, W., 2004a. Formation of secondary
organic aerosols through photooxidation of isoprene. Science (80-. ). 303, 1173–1176.
Claeys, M., Wang, W., Ion, A.C., Kourtchev, I., Gelencser, A., Maenhaut, W., 2004b.
Formation of secondary organic aerosols from isoprene and its gas-phase oxidation
products through reaction with hydrogen peroxide. Atmos. Environ. 38, 4093–4098.
Claeys, M., Wang, W., Vermeylen, R., Kourtchev, I., Chi, X.G., Farhat, Y., Surratt, J.D.,
Gomez-Gonzalez, Y., Sciare, J., Maenhaut, W., 2010. Chemical characterisation of
marine aerosol at Amsterdam Island during the austral summer of 2006-2007. J.
Aerosol Sci. 41, 13–22.
Coates, J., 2006. Interpretation of Infrared Spectra, A Practical Approach. In: Encyclopedia
of Analytical Chemistry. John Wiley & Sons, Ltd.
Coble, P.G., 1996. Characterization of marine and terrestrial DOM in seawater using
excitation emission matrix spectroscopy. Mar. Chem. 51, 325–346.
Cook, R., 2004. Coupling NMR to NOM. Anal. Bioanal. Chem. 378, 1484–1503.
Coury, C., Dillner, A.M., 2008. A method to quantify organic functional groups and
inorganic compounds in ambient aerosols using attenuated total reflectance FTIR
spectroscopy and multivariate chemometric techniques. Atmos. Environ. 42, 5923–
5932.
Coury, C., Dillner, A.M., 2009. ATR-FTIR characterization of organic functional groups
and inorganic ions in ambient aerosols at a rural site. Atmos. Environ. 43, 940–948.
DeCarlo, P.F., Kimmel, J.R., Trimborn, A., Northway, M.J., Jayne, J.T., Aiken, A. l,
Gonin, M., Fuhrer, K., Horvath, T., Docherty, K.S., Worsnop, D.R., Jimenez, J.L.,
2006. Field-Deployable, High-Resolution, Time-of-Flight Aerosol Mass
Spectrometer. Anal. Chem. 78, 8281–8289.
References
154
Decesari, S., Facchini, M.C., Fuzzi, S., McFiggans, G.B., Coe, H., Bower, K.N., 2005. The
water-soluble organic component of size-segregated aerosol, cloud water and wet
depositions from Jeju Island during ACE-Asia. Atmos. Environ. 39, 211–222.
Decesari, S., Facchini, M.C., Fuzzi, S., Tagliavini, E., 2000. Characterization of water-
soluble organic compounds in atmospheric aerosol: A new approach. J. Geophys. Res.
105, 1481–1489.
Decesari, S., Facchini, M.C., Matta, E., Lettini, F., Mircea, M., Fuzzi, S., Tagliavini, E.,
Putaud, J.P., 2001. Chemical features and seasonal variation of fine aerosol water-
soluble organic compounds in the Po Valley, Italy. Atmos. Environ. 35, 3691–3699.
Decesari, S., Fuzzi, S., Facchini, M.C., Mircea, M., Emblico, L., Cavalli, F., Maenhaut,
W., Chi, X., Schkolnik, G., Falkovich, A., Rudich, Y., Claeys, M., Pashynska, V.,
Vas, G., Kourtchev, I., Vermeylen, R., Hoffer, A., Andreae, M.O., Tagliavini, E.,
Moretti, F., Artaxo, P., 2006. Characterization of the organic composition of aerosols
from Rondonia, Brazil, during the LBA-SMOCC 2002 experiment and its
representation through model compounds. Atmos. Chem. Phys. 6, 375–402.
Decesari, S., Mircea, M., Cavalli, F., Fuzzi, S., Moretti, F., Tagliavini, E., Facchini, M.C.,
2007. Source attribution of water-soluble organic aerosol by nuclear magnetic
resonance spectroscopy. Environ. Sci. Technol. 41, 2479–2484.
Dinar, E., Mentel, T.F., Rudich, Y., 2006. The density of humic acids and humic like
substances (HULIS) from fresh and aged wood burning and pollution aerosol
particles. Atmos. Chem. Phys. 6, 5213–5224.
Dinar, E., Taraniuk, I., Graber, E.R., Anttila, T., Mentel, T.F., Rudich, Y., 2007.
Hygroscopic growth of atmospheric and model humic-like substances. J. Geophys.
Res. Atmos. 112, D05211.
Draxler, R.R., Rolph, G.D., 2013. HYSPLIT (HYbrid Single-Particle Lagrangian
Integrated Trajectory) Model access via NOAA ARL READY Website. NOAA Air
Resour. Lab.
References
155
Duarte, A.C., Duarte, R.M.B.O., 2009. Natural Organic Matter in Atmospheric Particles.
In: Senesi, N., Xing, B., Huang, P.M. (Eds.), Biophysico-Chemical Processes
Involving Natural Nonliving Organic Matter in Environmental Systems. John Wiley
& Sons, Inc., New Jersey, pp. 451–485.
Duarte, R.M.B.O., 2006. Balanço mássico e caracterização da matéria orgânica de
aerossóis atmosféricos. Tese de Doutoramento. Dep. Química. Universidade de
Aveiro.
Duarte, R.M.B.O., Barros, A.C., Duarte, A.C., 2012. Resolving the chemical heterogeneity
of natural organic matter: New insights from comprehensive two-dimensional liquid
chromatography. J. Chromatogr. A 1249, 138–146.
Duarte, R.M.B.O., Duarte, A.C., 2005. Application of non-ionic solid sorbents (XAD
resins) for the isolation and fractionation of water-soluble organic compounds from
atmospheric aerosols. J. Atmos. Chem. 51, 79–93.
Duarte, R.M.B.O., Duarte, A.C., 2011a. A critical review of advanced analytical
techniques for water-soluble organic matter from atmospheric aerosols. TrAC Trends
Anal. Chem. 30, 1659–1671.
Duarte, R.M.B.O., Duarte, A.C., 2011b. Optimizing size-exclusion chromatographic
conditions using a composite objective function and chemometric tools: Application
to natural organic matter profiling. Anal. Chim. Acta 688, 90–98.
Duarte, R.M.B.O., Duarte, A.C., 2013. Atmospheric Organic Matter. In: eMagRes. John
Wiley & Sons, Ltd.
Duarte, R.M.B.O., Fernández-Getino, A.P., Duarte, A.C., 2013. Humic acids as proxies for
assessing different Mediterranean forest soils signatures using solid-state CPMAS
13C NMR spectroscopy. Chemosphere 91, 1556–1565.
Duarte, R.M.B.O., Mieiro, C.L., Penetra, A., Pio, C.A., Duarte, A.C., 2008a. Carbonaceous
materials in size-segregated atmospheric aerosols from urban and coastal-rural areas
at the Western European Coast. Atmos. Res. 90, 253–263.
References
156
Duarte, R.M.B.O., Pio, C.A., Duarte, A.C., 2004. Synchronous scan and excitation-
emission matrix fluorescence spectroscopy of water-soluble organic compounds in
atmospheric aerosols. J. Atmos. Chem. 48, 157–171.
Duarte, R.M.B.O., Pio, C.A., Duarte, A.C., 2005. Spectroscopic study of the water-soluble
organic matter isolated from atmospheric aerosols collected under different
atmospheric conditions. Anal. Chim. Acta 530, 7–14.
Duarte, R.M.B.O., Santos, E.B.H., Duarte, A.C., 2003. Spectroscopic characteristics of
ultrafiltration fractions of fulvic and humic acids isolated from an eucalyptus bleached
Kraft pulp mill effluent. Water Res. 37, 4073–4080.
Duarte, R.M.B.O., Santos, E.B.H., Pio, C.A., Duarte, A.C., 2007. Comparison of structural
features of water-soluble organic matter from atmospheric aerosols with those of
aquatic humic substances. Atmos. Environ. 41, 8100–8113.
Duarte, R.M.B.O., Silva, A.M.S., Duarte, A.C., 2008b. Two-Dimensional NMR Studies of
Water-Soluble Organic Matter in Atmospheric Aerosols. Environ. Sci. Technol. 42,
8224–8230.
Dugo, P., Cacciola, F., Donato, P., Airado-Rodríguez, D., Herrero, M., Mondello, L., 2009.
Comprehensive two-dimensional liquid chromatography to quantify polyphenols in
red wines. J. Chromatogr. A 1216, 7483–7487.
Dugo, P., Cacciola, F., Kumm, T., Dugo, G., Mondello, L., 2008. Comprehensive
multidimensional liquid chromatography: Theory and applications. J. Chromatogr. A
1184, 353–368.
Dugo, P., Kumm, T., Crupi, M.L., Cotroneo, A., Mondello, L., 2006. Comprehensive two-
dimensional liquid chromatography combined with mass spectrometric detection in
the analyses of triacylglycerols in natural lipidic matrixes. J. Chromatogr. A 1112,
269–275.
Facchini, M.C., Fuzzi, S., Zappoli, S., Andracchio, A., Gelencser, A., Kiss, G., Krivacsy,
Z., Meszaros, E., Hansson, H.C., Alsberg, T., Zebuhr, Y., 1999. Partitioning of the
References
157
organic aerosol component between fog droplets and interstitial air. J. Geophys. Res.
104, 26821–26832.
Facchini, M.C., Rinaldi, M., Decesari, S., Carbone, C., Finessi, E., Mircea, M., Fuzzi, S.,
Ceburnis, D., Flanagan, R., Nilsson, E.D., Leeuw, G., Martino, M., Woeltjen, J.,
O’Dowd, C.D., 2008. Primary submicron marine aerosol dominated by insoluble
organic colloids and aggregates. Geophys. Res. Lett. 35, L17814.
Faix, O., 1992. No Title. In: Lin, S.Y., Dence, C.W. (Eds.), Methods in Lignin Chemistry.
Springer, Berlin, pp. 92–93.
Feng, J.L., Chan, C.K., Fang, M., Hu, M., He, L.Y., Tang, X.Y., 2006. Characteristics of
organic matter in PM2.5 in Shanghai. Chemosphere 64, 1393–1400.
Fengel, D., Wegner, G., 1984. Wood: chemistry, ultrastructure, reactions. W. de Gruyter,
Berlin.
Fuzzi, S., Andreae, M.O., Huebert, B.J., Kulmala, M., Bond, T.C., Boy, M., Doherty, S.J.,
Guenther, A., Kanakidou, M., Kawamura, K., Kerminen, V.-M., Lohmann, U.,
Russell, L.M., Pöschl, U., 2006. Critical assessment of the current state of scientific
knowledge, terminology, and research needs concerning the role of organic aerosols
in the atmosphere, climate, and global change. Atmos. Chem. Phys. 6, 2017–2038.
Fuzzi, S., Decesari, S., Facchini, M.C., Matta, E., Mircea, M., Tagliavini, E., 2001. A
simplified model of the water soluble organic component of atmospheric aerosols.
Geophys. Res. Lett. 28, 4079–4082.
Gelencser, A., 2004. Carbonaceous Aerosol. In: Atmospheric and Oceanographic Sciences
Library. Springer, Dordrecht.
Gelencser, A., Hoffer, A., Krivacsy, Z., Kiss, G., Molnar, A., Meszaros, E., 2002. On the
possible origin of humic matter in fine continental aerosol. J. Geophys. Res. 107.
References
158
Giovanela, M., Parlanti, E., Soriano-Sierra, E.J., Soldi, M.S., Sierra, M.M.D., 2004.
Elemental compositions, FT-IR spectra and thermal behavior of sedimentary fulvic
and humic acids from aquatic and terrestrial environments. Geochem. J. 38, 255–264.
Góra, R., Hutta, M., Rohárik, P., 2012. Characterization and analysis of soil humic acids
by off-line combination of wide-pore octadecylsilica column reverse phase high
performance liquid chromatography with narrow bore column size-exclusion
chromatography and fluorescence detection. J. Chromatogr. A 1220, 44–49.
Graber, E.R., Rudich, Y., 2006. Atmospheric HULIS: How humic-like are they? A
comprehensive and critical review. Atmos. Chem. Phys. 6, 729–753.
Graham, B., Mayol-Bracero, O.L., Guyon, P., Roberts, G.C., Decesari, S., Facchini, M.C.,
Artaxo, P., Maenhaut, W., Koll, P., Andreae, M.O., 2002. Water-soluble organic
compounds in biomass burning aerosols over Amazonia - 1. Characterization by
NMR and GC-MS. J. Geophys. Res. 107.
Gysel, M., Weingartner, E., Nyeki, S., Paulsen, D., Baltensperger, U., Galambos, I., Kiss,
G., 2004. Hygroscopic properties of water-soluble matter and humic-like organics in
atmospheric fine aerosol. Atmos. Chem. Phys. 4, 35–50.
Hatfield, G.R., Maciel, G.E., Erbatur, O., Erbatur, G., 1987. Qualitative and quantitative
analysis of solid lignin samples by carbon-13 nuclear magnetic resonance
spectrometry. Anal. Chem. 59, 172–179.
Havers, N., Burba, P., Lambert, J., Klockow, D., 1998. Spectroscopic characterization of
humic-like substances in airborne particulate matter. J. Atmos. Chem. 29, 45–54.
Haw, J.F., Maciel, G.E., Schroeder, H.A., 1984. Carbon-13 nuclear magnetic resonance
spectrometric study of wood and wood pulping with cross polarization and magic-
angle spinning. Anal. Chem. 56, 1323–1329.
Hawkins, L.N., Russell, L.M., 2010. Oxidation of ketone groups in transported biomass
burning aerosol from the 2008 Northern California Lightning Series fires. Atmos.
Environ. 44, 4142–4154.
References
159
Hayes, M.H.B., MacCarthy, P., Malcolm, R.L., Swift, R.S., 1989. Humic substances II. In
search of structure. John Wiley & Sons Ltd.
Ho, K.F., Lee, S.C., Chan, C.K., Yu, J.C., Chow, J.C., Yao, X.H., 2003. Characterization
of chemical species in PM2.5 and PM10 aerosols in Hong Kong. Atmos. Environ. 37,
31–39.
Ho, K.F., Lee, S.C., Yu, J.C., Zou, S.C., Fung, K., 2002. Carbonaceous characteristics of
atmospheric particulate matter in Hong Kong. Sci. Total Environ. 300, 59–67.
Hoffer, A., Gelencser, A., Guyon, P., Kiss, G., Schmid, O., Frank, G.P., Artaxo, P.,
Andreae, M.O., 2006. Optical properties of humic-like substances (HULIS) in
biomass-burning aerosols. Atmos. Chem. Phys. 6, 3563–3570.
Huang, X., Yu, J.Z., He, L., Zibing, Z., 2006. Water-soluble organic carbon and oxalate in
aerosols at a coastal urban site in China: Size distribution characteristics, sources, and
formation mechanisms. J. Geophys. Res. Atmos. 111, D22212.
IPCC, 2001. Aerosols, their direct and indirect effects. Climate Change 2001: The
Scientific Basis. Contribution of Working Group I to the Third Assessment Report of
the Intergovernmental Panel on Climate Change.
Jacobson, M.C., Hansson, H.C., Noone, K.J., Charlson, R.J., 2000. Organic atmospheric
aerosols: Review and state of the science. Rev. Geophys. 38, 267–294.
Jaffrezo, J.L., Aymoz, G., Delaval, C., Cozic, J., 2005. Seasonal variations of the water
soluble organic carbon mass fraction of aerosol in two valleys of the French Alps.
Atmos. Chem. Phys. 5, 2809–2821.
Kanakidou, M., Seinfeld, J.H., Pandis, S.N., Barnes, I., Dentener, F.J., Facchini, M.C., Van
Dingenen, R., Ervens, B., Nenes, A., Nielsen, C.J., Swietlicki, E., Putaud, J.P.,
Balkanski, Y., Fuzzi, S., Horth, J., Moortgat, G.K., Winterhalter, R., Myhre, C.E.L.,
Tsigaridis, K., Vignati, E., Stephanou, E.G., Wilson, J., 2005. Organic aerosol and
global climate modelling: a review. Atmos. Chem. Phys. 5, 1053–1123.
References
160
Kaufman, Y.J., Tanre, D., Boucher, O., 2002. A satellite view of aerosols in the climate
system. Nature 419, 215–223.
Keeler, C., Kelly, E.F., Maciel, G.E., 2006. Chemical–structural information from solid-
state 13C NMR studies of a suite of humic materials from a lower montane forest soil,
Colorado, USA. Geoderma 130, 124–140.
Khare, P., Baruah, B.P., Rao, P.G., 2011. Water-soluble organic compounds (WSOCs) in
PM2.5 and PM10 at a subtropical site of India. Tellus Ser. B-Chemical Phys.
Meteorol. 63, 990–1000.
Kim, K.H., Sekiguchi, K., Furuuchi, M., Sakamoto, K., 2011. Seasonal variation of
carbonaceous and ionic components in ultrafine and fine particles in an urban area of
Japan. Atmos. Environ. 45, 1581–1590.
Kiss, G., Gelencser, A., Hoffer, A., Krivacsy, Z., Meszaros, E., Molnar, A., Varga, B.,
2000. Chemical characterization of water soluble organic compounds in tropospheric
fine aerosol. Nucleation Atmos. Aerosols 534, 761–764.
Kiss, G., Tombacz, E., Varga, B., Alsberg, T., Persson, L., 2003. Estimation of the average
molecular weight of humic-like substances isolated from fine atmospheric aerosol.
Atmos. Environ. 37, 3783–3794.
Kiss, G., Varga, B., Galambos, I., Ganszky, I., 2002. Characterization of water-soluble
organic matter isolated from atmospheric fine aerosol. J. Geophys. Res. 107, Doi
10.1029/2001jd000603.
Kivilompolo, M., Hyötyläinen, T., 2007. Comprehensive two-dimensional liquid
chromatography in analysis of Lamiaceae herbs: Characterisation and quantification
of antioxidant phenolic acids. J. Chromatogr. A 1145, 155–164.
Kleindienst, T.E., Jaoui, M., Lewandowski, M., Offenberg, J.H., Lewis, C.W., Bhave, P.
V, Edney, E.O., 2007. Estimates of the contributions of biogenic and anthropogenic
hydrocarbons to secondary organic aerosol at a southeastern US location. Atmos.
Environ. 41, 8288–8300.
References
161
Kondo, Y., Miyazaki, Y., Takegawa, N., Miyakawa, T., Weber, R.J., Jimenez, J.L., Zhang,
Q., Worsnop, D.R., 2007. Oxygenated and water-soluble organic aerosols in Tokyo. J.
Geophys. Res. 112.
Krivacsy, Z., Gelencser, A., Kiss, G., Meszaros, E., Molnar, A., Hoffer, A., Meszaros, T.,
Sarvari, Z., Temesi, D., Varga, B., Baltensperger, U., Nyeki, S., Weingartner, E.,
2001a. Study on the chemical character of water soluble organic compounds in fine
atmospheric aerosol at the Jungfraujoch. J. Atmos. Chem. 39, 235–259.
Krivacsy, Z., Hoffer, A., Sarvari, Z., Temesi, D., Baltensperger, U., Nyeki, S.,
Weingartner, E., Kleefeld, S., Jennings, S.G., 2001b. Role of organic and black
carbon in the chemical composition of atmospheric aerosol at European background
sites. Atmos. Environ. 35, 6231–6244.
Krivacsy, Z., Kiss, G., Ceburnis, D., Jennings, G., Maenhaut, W., Salma, I., Shooter, D.,
2008. Study of water-soluble atmospheric humic matter in urban and marine
environments. Atmos. Res. 87, 1–12.
Krivacsy, Z., Kiss, G., Varga, B., Galambos, I., Sarvari, Z., Gelencser, A., Molnar, A.,
Fuzzi, S., Facchini, M.C., Zappoli, S., Andracchio, A., Alsberg, T., Hansson, H.C.,
Persson, L., 2000. Study of humic-like substances in fog and interstitial aerosol by
size-exclusion chromatography and capillary electrophoresis. Atmos. Environ. 34,
4273–4281.
Krivácsy, Z., Molnár, Á., 1998. Size distribution of ions in atmospheric aerosols. Atmos.
Res. 46, 279–291.
Lambert, J., Lankes, U., 2002. Structural Investigations: Application of Nuclear Magnetic
Resonance Spectroscopy to Structural Investigations of Refractory Organic
Substances – Principles and Definitions. In: Frimmel, F.H., Abbt-Braun, G.,
Heumann, K.G., Hock, B., Lüdemann, H.-D., Spiteller, M. (Eds.), Refractory Organic
Substances in the Environment. Wiley-VCH, pp. 89–95.
References
162
Limbeck, A., Kulmala, M., Puxbaum, H., 2003. Secondary organic aerosol formation in
the atmosphere via heterogeneous reaction of gaseous isoprene on acidic particles.
Geophys. Res. Lett. 30, n/a–n/a.
Lin, Y.C., Cheng, M.T., Chio, C.P., Kuo, C.Y., 2009. Carbonaceous Aerosol
Measurements at Coastal, Urban, and Inland Sites in Central Taiwan. Environ.
Forensics 10, 7–17.
MADRP-Direcção de Unidade de Defesa da Floresta, 2011. Relatório anual de áreas
ardidas e ocorrências 2010. Portugal.
Maimone, F., Turpin, B.J., Solomon, P., Meng, Q.Y., Robinson, A.L., Subramanian, R.,
Polidori, A., 2011. Correction Methods for Organic Carbon Artifacts When Using
Quartz-Fiber Filters in Large Particulate Matter Monitoring Networks: The
Regression Method and Other Options. J. Air Waste Manage. Assoc. 61, 696–710.
Malcolm, R.L., 1989. Applications of solid-state 13C-NMR spectroscopy to geochemical
studies of humic substances. In: Hayes, M.H.B., MacCarthy, P., Malcolm, R.L.,
Swift, R.S. (Eds.), Humic Substances II: In Search of Structure. Wiley, Chichester,
pp. 340–372.
Maria, S.F., Russell, L.M., Turpin, B.J., Porcja, R.J., 2002. FTIR measurements of
functional groups and organic mass in aerosol samples over the Caribbean. Atmos.
Environ. 36, 5185–5196.
Matos, J.T. V, Duarte, R.M.B.O., Duarte, A.C., 2012. Trends in data processing of
comprehensive two-dimensional chromatography: State of the art. J. Chromatogr. B
910, 31–45.
Matos, J.T. V, Duarte, R.M.B.O., Duarte, A.C., 2013. A simple approach to reduce
dimensionality from comprehensive two-dimensional liquid chromatography coupled
with a multichannel detector. Anal. Chim. Acta 804, 296–303.
Mayol-Bracero, O.L., Guyon, P., Graham, B., Roberts, G., Andreae, M.O., Decesari, S.,
Facchini, M.C., Fuzzi, S., Artaxo, P., 2002. Water-soluble organic compounds in
References
163
biomass burning aerosols over Amazonia - 2. Apportionment of the chemical
composition and importance of the polyacidic fraction. J. Geophys. Res. 107, Doi
10.1029/2001jd000522.
Mazzoleni, L.R., Ehrmann, B.M., Shen, X.H., Marshall, A.G., Collett, J.L., 2010. Water-
Soluble Atmospheric Organic Matter in Fog: Exact Masses and Chemical Formula
Identification by Ultrahigh-Resolution Fourier Transform Ion Cyclotron Resonance
Mass Spectrometry. Environ. Sci. Technol. 44, 3690–3697.
McDonald, S., Bishop, A.G., Prenzler, P.D., Robards, K., 2004. Analytical chemistry of
freshwater humic substances. Anal. Chim. Acta 527, 105–124.
McKnight, D.M., Boyer, E.W., Westerhoff, P.K., Doran, P.T., Kulbe, T., Andersen, D.T.,
2001. Spectrofluorometric characterization of dissolved organic matter for indication
of precursor organic material and aromaticity. Limnol. Oceanogr. 46, 38–48.
McMurry, P.H., 2000. A review of atmospheric aerosol measurements. Atmos. Environ.
34, 1959–1999.
Miller, J.N., Miller, J.C., 2005. Statistics and Chemometrics for Analytical Chemistry,
Fifth. ed. England: Pearson - Prentice Hall.
Mircea, M., Facchini, M.C., Decesari, S., Fuzzi, S., Charlson, R.J., 2002. The influence of
the organic aerosol component on CCN supersaturation spectra for different aerosol
types. Tellus B 54, 74–81.
Miyazaki, Y., Kondo, Y., Takegawa, N., Komazaki, Y., Fukuda, M., Kawamura, K.,
Mochida, M., Okuzawa, K., Weber, R.J., 2006. Time-resolved measurements of
water-soluble organic carbon in Tokyo. J. Geophys. Res. 111.
Mkoma, S.L., Chi, X.G., Maenhaut, W., 2010. Characteristics of carbonaceous aerosols in
ambient PM10 and PM2.5 particles in Dar es Salaam, Tanzania. Sci. Total Environ.
408, 1308–1314.
References
164
Mladenov, N., Alados-Arboledas, L., Olmo, F.J., Lyamani, H., Delgado, A., Molina, A.,
Reche, I., 2011. Applications of optical spectroscopy and stable isotope analyses to
organic aerosol source discrimination in an urban area. Atmos. Environ. 45, 1960–
1969.
Mladenov, N., Reche, I., Olmo, F.J., Lyamani, H., Alados-Arboledas, L., 2010.
Relationships between spectroscopic properties of high-altitude organic aerosols and
Sun photometry from ground-based remote sensing. J. Geophys. Res. Biogeosciences
115, G00F11.
Nakajima, H., Okada, K., Kuroki, Y., Nakama, Y., Handa, D., Arakaki, T., Tanahara, A.,
2008. Photochemical formation of peroxides and fluorescence characteristics of the
water-soluble fraction of bulk aerosols collected in Okinawa, Japan. Atmos. Environ.
42, 3046–3058.
Navrátil, T., Hladil, J., Strnad, L., Koptíková, L., Skála, R., 2012. Volcanic ash particulate
matter from the 2010 Eyjafjallajökull eruption in dust deposition at Prague, central
Europe. Aeolian Res.
Park, S.S., Cho, S.Y., 2011. Tracking sources and behaviors of water-soluble organic
carbon in fine particulate matter measured at an urban site in Korea. Atmos. Environ.
45, 60–72.
Parlanti, E., Wörz, K., Geoffroy, L., Lamotte, M., 2000. Dissolved organic matter
fluorescence spectroscopy as a tool to estimate biological activity in a coastal zone
submitted to anthropogenic inputs. Org. Geochem. 31, 1765–1781.
Peltier, R.E., Weber, R.J., Sullivan, A.P., 2007. Investigating a liquid-based method for
online organic carbon detection in atmospheric particles. Aerosol Sci. Technol. 41,
1117–1127.
Pereira, A.S., David, F., Vanhoenacker, G., Sandra, P., 2009. The acetonitrile shortage: Is
reversed HILIC with water an alternative for the analysis of highly polar ionizable
solutes? J. Sep. Sci. 32, 2001–2007.
References
165
Peuravuori, J., Koivikko, R., Pihlaja, K., 2002a. Characterization. differentiation and
classification of aquatic humic matter separated with different sorbents: synchronous
scanning fluorescence spectroscopy. Water Res. 36, 4552–4562.
Peuravuori, J., Lehtonen, T., Pihlaja, K., 2002b. Sorption of aquatic humic matter by
DAX-8 and XAD-8 resins: Comparative study using pyrolysis gas chromatography.
Anal. Chim. Acta 471, 219–226.
Peuravuori, J., Pihlaja, K., 1997. Molecular size distribution and spectroscopic properties
of aquatic humic substances. Anal. Chim. Acta 337, 133–149.
Pio, C.A., Alves, C.A., Duarte, A.C., 2001. Identification, abundance and origin of
atmospheric organic particulate matter in a Portuguese rural area. Atmos. Environ. 35,
1365–1375.
Pio, C.A., Legrand, M., Alves, C.A., Oliveira, T., Afonso, J., Caseiro, A., Puxbaum, H.,
Sanchez-Ochoa, A., Gelencsér, A., 2008. Chemical composition of atmospheric
aerosols during the 2003 summer intense forest fire period. Atmos. Environ. 42,
7530–7543.
Pól, J., Hohnová, B., Jussila, M., Hyötyläinen, T., 2006. Comprehensive two-dimensional
liquid chromatography–time-of-flight mass spectrometry in the analysis of acidic
compounds in atmospheric aerosols. J. Chromatogr. A 1130, 64–71.
Poschl, U., 2005. Atmospheric aerosols: Composition, transformation, climate and health
effects. Angew. Chemie-International Ed. 44, 7520–7540.
Putaud, J.-P., Van Dingenen, R., Alastuey, A., Bauer, H., Birmili, W., Cyrys, J., Flentje,
H., Fuzzi, S., Gehrig, R., Hansson, H.C., Harrison, R.M., Herrmann, H., Hitzenberger,
R., Hüglin, C., Jones, A.M., Kasper-Giebl, A., Kiss, G., Kousa, A., Kuhlbusch,
T.A.J., Löschau, G., Maenhaut, W., Molnar, A., Moreno, T., Pekkanen, J., Perrino, C.,
Pitz, M., Puxbaum, H., Querol, X., Rodriguez, S., Salma, I., Schwarz, J., Smolik, J.,
Schneider, J., Spindler, G., ten Brink, H., Tursic, J., Viana, M., Wiedensohler, A.,
Raes, F., 2010. A European aerosol phenomenology – 3: Physical and chemical
References
166
characteristics of particulate matter from 60 rural, urban, and kerbside sites across
Europe. Atmos. Environ. 44, 1308–1320.
Ram, K., Sarin, M.M., 2010. Spatio-temporal variability in atmospheric abundances of EC,
OC and WSOC over Northern India. J. Aerosol Sci. 41, 88–98.
Reemtsma, T., 2009. Determination of molecular formulas of natural organic matter
molecules by (ultra-) high-resolution mass spectrometry Status and needs. J.
Chromatogr. A 1216, 3687–3701.
Remoundaki, E., Papayannis, A., Kassomenos, P., Mantas, E., Kokkalis, P., Tsezos, M.,
2012. Influence of Saharan Dust Transport Events on PM2.5 Concentrations and
Composition over Athens. Water, Air, Soil Pollut. 224, 1–14.
Rogge, W.F., Hildemann, L.M., Mazurek, M.A., Cass, G.R., Simoneit, B.R.T., 1998.
Sources of fine organic aerosol. 9. Pine, oak and synthetic log combustion in
residential fireplaces. Environ. Sci. Technol. 32, 13–22.
Saarikoski, S., Sillanpaa, M., Sofiev, M., Timonen, H., Saarnio, K., Teinela, K.,
Karppinen, A., Kukkonen, J., Hillamo, R., 2007. Chemical composition of aerosols
during a major biomass burning episode over northern Europe in spring 2006:
Experimental and modelling assessments. Atmos. Environ. 41, 3577–3589.
Saarikoski, S., Timonen, H., Saarnio, K., Aurela, M., Jarvi, L., Keronen, P., Kerminen,
V.M., Hillamo, R., 2008. Sources of organic carbon in fine particulate matter in
northern European urban air. Atmos. Chem. Phys. 8, 6281–6295.
Salma, I., Ocskay, R., Chi, X.G., Maenhaut, W., 2007. Sampling artefacts, concentration
and chemical composition of fine water-soluble organic carbon and humic-like
substances in a continental urban atmospheric environment. Atmos. Environ. 41,
4106–4118.
Salma, I., Ocskay, R., Varga, I., Maenhaut, W., 2006. Surface tension of atmospheric
humic-like substances in connection with relaxation, dilution, and solution pH. J.
Geophys. Res. 111, -.
References
167
Samara, C., Voutsa, D., 2005. Size distribution of airborne particulate matter and
associated heavy metals in the roadside environment. Chemosphere 59, 1197–1206.
Samburova, V., Zenobi, R., Kalberer, M., 2005. Characterization of high molecular weight
compounds in urban atmospheric particles. Atmos. Chem. Phys. 5, 2163–2170.
Sannigrahi, P., Sullivan, A.P., Weber, R.J., Ingall, E.D., 2006. Characterization of water-
soluble organic carbon in urban atmospheric aerosols using solid-state C-13 NMR
spectroscopy. Environ. Sci. Technol. 40, 666–672.
Santos, E.B.H., Duarte, A.C., 1998. The influence of pulp and paper mill effluents on the
composition of the humic fraction of aquatic organic matter. Water Res. 32, 597–608.
Santos, E.B.H., Duarte, R.M.B.O., Filipe, O.S., Duarte, A.C., 2000. Structural
characterisation of the coloured organic matter from an eucalyptus pleached Kraft
pulp mill effluent. Int. J. Environ. Anal. Chem. 78, 333–342.
Santos, E.B.H., Filipe, O.M.S., Duarte, R.M.B.O., Pinto, H., Duarte, A.C., 2001.
Fluorescence as a tool for tracing the organic contamination from pulp mill effluents
in surface waters. Acta Hydrochim. Hydrobiol. 28, 364–371.
Santos, P.S.M., Duarte, R.M.B.O., Duarte, A.C., 2009. Absorption and fluorescence
properties of rainwater during the cold season at a town in Western Portugal. J.
Atmos. Chem. 62, 45–57.
Schmitt- opplin, P., elencs r, A., Dabek- lotorzynska, E., iss, ., Hertkorn, N., Harir,
M., Hong, ., ebef gi, I., 2010. Analysis of the Unresolved Organic Fraction in
Atmospheric Aerosols with Ultrahigh-Resolution Mass Spectrometry and Nuclear
Magnetic Resonance Spectroscopy: Organosulfates As Photochemical Smog
Constituents†. Anal. Chem. 82, 8017–8026.
Seinfeld, J.H., 1986. Atmospheric chemistry and physics of air pollution. John Wiley &
Sons, Inc., New York, EUA.
References
168
Seinfeld, J.H., Pandis, S.N., 1998. Atmospheric Chemistry and Physics: From Air
Pollution to Climate Change. John Wiley & Sons, Inc., New York, EUA.
Seinfeld, J.H., Pandis, S.N., 2006. Atmospheric Chemistry and Physics: From Air
Pollution to Climate Change., 2nd ed. John Wiley & Sons, Canada.
Seinfeld, J.H., Pankow, J.F., 2003. Organic atmospheric particulate material. Annu. Rev.
Phys. Chem. 54, 121–140.
Sempére, R., Kawamura, K., 1994. Comparative distributions of dicarboxylic acids and
related polar compounds in snow, rain and aerosols from urban atmosphere. Atmos.
Environ. 28, 449–459.
Senesi, N., Miano, T.M., Provenzano, M.R., Brunetti, G., 1991. Characterization,
differentiation, and classification of humic substances by fluorescence spectroscopy.
Soil Sci. 152, 259–271.
Simoneit, B.R.T., Rogge, W.F., Mazurek, M.A., Standley, L.J., Hildemann, L.M., Cass,
G.R., 1993. Lignin pyrolysis products, lignans, and resin acids as specific tracers of
plant classes in emissions from biomass combustion. Environ. Sci. Technol. 27,
2533–2541.
Simoneit, B.R.T., Schauer, J.J., Nolte, C.G., Oros, D.R., Elias, V.O., Fraser, M.P., Rogge,
W.F., Cass, G.R., 1999. Levoglucosan, a tracer for cellulose in biomass burning and
atmospheric particles. Atmos. Environ. 33, 173–182.
Simpson, A.J., 2001. Multidimensional solution state NMR of humic substances: A
practical guide and review. Soil Sci. 166, 795–809.
Simpson, A.J., Simpson, M.J., 2009. Nuclear Magnetic Resonance Analysis of Natural
Organic Matter. In: Senesi, N., Xing, B., Huang, P.. (Eds.), Biophysico-Chemical
Processes Involving Natural Nonliving Organic Matter in Environmental Systems.
John Wiley & Sons, Inc., New Jersey, pp. 589–650.
References
169
Smith, D.M., Chughtai, A.R., 1995. The surface structure and reactivity of black carbon.
Colloids Surfaces A Physicochem. Eng. Asp. 105, 47–77.
Snyder, D.C., Rutter, A.P., Collins, R., Worley, C., Schauer, J.J., 2009. Insights into the
Origin of Water Soluble Organic Carbon in Atmospheric Fine Particulate Matter.
Aerosol Sci. Technol. 42, 1099–1107.
Stoll, D.R., Li, X., Wang, X., Carr, P.W., Porter, S.E.G., Rutan, S.C., 2007. Fast,
comprehensive two-dimensional liquid chromatography. J. Chromatogr. A 1168, 3–
43.
Stone, E.A., Hedman, C.J., Sheesley, R.J., Shafer, M.M., Schauer, J.J., 2009. Investigating
the chemical nature of humic-like substances (HULIS) in North American
atmospheric aerosols by liquid chromatography tandem mass spectrometry. Atmos.
Environ. 43, 4205–4213.
Subbalakshmi, Y., Patti, A.F., Lee, G.S.H., Hooper, M.A., 2000. Structural
characterisation of macromolecular organic material in air particulate matter using Py-
GC-MS and solid state C-13-NMR. J. Environ. Monit. 2, 561–565.
Subramanian, R., Khlystov, A.Y., Cabada, J.C., Robinson, A.L., 2004. Positive and
Negative Artifacts in Particulate Organic Carbon Measurements with Denuded and
Undenuded Sampler Configurations Special Issue of Aerosol Science and Technology
on Findings from the Fine Particulate Matter Supersites Program. Aerosol Sci.
Technol. 38, 27–48.
Sullivan, A.P., Weber, R.J., 2006. Chemical characterization of the ambient organic
aerosol soluble in water: 1. Isolation of hydrophobic and hydrophilic fractions with a
XAD-8 resin - art. no. D05314. J Geophys Res Atmos 111, 5314.
Sullivan, A.P., Weber, R.J., Clements, A.L., Turner, J.R., Bae, M.S., Schauer, J.J., 2004. A
method for on-line measurement of water-soluble organic carbon in ambient aerosol
particles: Results from an urban site. Geophys. Res. Lett. 31, -.
References
170
Sun, Y., Zhang, Q., Zheng, M., Ding, X., Edgerton, E.S., Wang, X., 2011. Characterization
and Source Apportionment of Water-Soluble Organic Matter in Atmospheric Fine
Particles (PM2.5) with High-Resolution Aerosol Mass Spectrometry and GC–MS.
Environ. Sci. Technol. 45, 4854–4861.
Takahama, S., Schwartz, R.E., Russell, L.M., Macdonald, A.M., Sharma, S., Leaitch,
W.R., 2011. Organic functional groups in aerosol particles from burning and non-
burning forest emissions at a high-elevation mountain site. Atmos. Chem. Phys. 11,
6367–6386.
Timonen, H., Aurela, M., Carbone, S., Saarnio, K., Saarikoski, S., Makela, T., Kulmala,
M., Kerminen, V.M., Worsnop, D.R., Hillamo, R., 2010. High time-resolution
chemical characterization of the water-soluble fraction of ambient aerosols with PILS-
TOC-IC and AMS. Atmos. Meas. Tech. 3, 1063–1074.
Timonen, H.J., Saarikoski, S.K., Aurela, M.A., Saarnio, K.M., Hillamo, R.E.J., 2008.
Water-soluble organic carbon in urban aerosol: concentrations, size distributions and
contribution to particulate matter. Boreal Environ. Res. 13, 335–346.
Turpin, B.J., Huntzicker, J.J., Hering, S. V, 1994. Investigation of organic aerosol
sampling artifacts in the los-angeles basin. Atmos. Environ. 28, 3061–3071.
Turpin, B.J., Lim, H.J., 2001. Species contributions to PM2.5 mass concentrations:
Revisiting common assumptions for estimating organic mass. Aerosol Sci. Technol.
35, 602–610.
Turpin, B.J., Saxena, P., Andrews, E., 2000. Measuring and simulating particulate organics
in the atmosphere: problems and prospects. Atmos. Environ. 34, 2983–3013.
Varga, B., Kiss, G., Ganszky, I., Gelencser, A., Krivacsy, Z., 2001. Isolation of water-
soluble organic matter from atmospheric aerosol. Talanta 55, 561–572.
Viana, M., Chi, X., Maenhaut, W., Cafmeyer, J., Querol, X., Alastuey, A., Mikuska, P.,
Vecera, Z., 2006a. Influence of sampling artefacts on measured PM, OC, and EC
levels in Carbonaceous aerosols in an urban area. Aerosol Sci. Technol. 40, 107–117.
References
171
Viana, M., Chi, X., Maenhaut, W., Querol, X., Alastuey, A., Mikuska, P., Vecera, Z.,
2006b. Organic and elemental carbon concentrations in carbonaceous aerosols during
summer and winter sampling campaigns in Barcelona, Spain. Atmos. Environ. 40,
2180–2193.
Viana, M., Maenhaut, W., ten Brink, H.M., Chi, X., Weijers, E., Querol, X., Alastuey, A.,
Mikuška, P., Večeřa, ., 2007. Comparative analysis of organic and elemental carbon
concentrations in carbonaceous aerosols in three European cities. Atmos. Environ. 41,
5972–5983.
Viidanoja, J., Sillanpaa, M., Laakia, J., Kerminen, V.M., Hillamo, R., Aarnio, P.,
Koskentalo, T., 2002. Organic and black carbon in PM2.5 and PM10: 1 year of data
from an urban site in Helsinki, Finland. Atmos. Environ. 36, 3183–3193.
Wang, Y., Chiu, C., Westerhoff, P., Valsaraj, K.T., Herckes, P., 2013. Characterization of
atmospheric organic matter using size-exclusion chromatography with inline organic
carbon detection. Atmos. Environ. 68, 326–332.
Weber, R.J., Orsini, D., Daun, Y., Lee, Y.-N., Klotz, P.J., Brechtel, F., 2001. A Particle-
into-Liquid Collector for Rapid Measurement of Aerosol Bulk Chemical
Composition. Aerosol Sci. Technol. 35, 718–727.
Weber, R.J., Sullivan, A.P., Peltier, R.E., Russell, A., Yan, B., Zheng, M., de Gouw, J.,
Warneke, C., Brock, C., Holloway, J.S., Atlas, E.L., Edgerton, E., 2007. A study of
secondary organic aerosol formation in the anthropogenic-influenced southeastern
United States. J. Geophys. Res. Atmos. 112, D13302.
Wozniak, A.S., Bauer, J.E., Sleighter, R.L., Dickhut, R.M., Hatcher, P.G., 2008. Technical
Note: Molecular characterization of aerosol-derived water soluble organic carbon
using ultrahigh resolution electrospray ionization Fourier transform ion cyclotron
resonance mass spectrometry. Atmos. Chem. Phys. 8, 5099–5111.
Yamashita, Y., Tanoue, E., 2003. Chemical characterization of protein-like fluorophores in
DOM in relation to aromatic amino acids. Mar. Chem. 82, 255–271.
References
172
Yang, H., Xu, J., Wu, W., Wan, C.H., Yu, J.Z., 2004. Chemical Characterization of Water-
Soluble Organic Aerosols at Jeju Island Collected During ACE-Asia. Environ. Chem.
1, 13–17.
Yoon, Y.J., Ceburnis, D., Cavalli, F., Jourdan, O., Putaud, J.P., Facchini, M.C., Decesari,
S., Fuzzi, S., Sellegri, ., Jennings, S. ., O’Dowd, C.D., 2007. Seasonal
characteristics of the physicochemical properties of North Atlantic marine
atmospheric aerosols. J. Geophys. Res. Atmos. 112, D04206.
Yu, L.E., Shulman, M.L., Kopperud, R., Hildemann, L.M., 2004. Characterization of
Organic Compounds Collected during Southeastern Aerosol and Visibility Study:
Water-Soluble Organic Species. Environ. Sci. Technol. 39, 707–715.
Zappoli, S., Andracchio, A., Fuzzi, S., Facchini, M.C., Gelencser, A., Kiss, G., Krivacsy,
Z., Molnar, A., Meszaros, E., Hansson, H.C., Rosman, K., Zebuhr, Y., 1999.
Inorganic, organic and macromolecular components of fine aerosol in different areas
of Europe in relation to their water solubility. Atmos. Environ. 33, 2733–2743.
Zhang, F., Zhao, J., Chen, J., Xu, Y., Xu, L., 2011. Pollution characteristics of organic and
elemental carbon in PM2.5 in Xiamen, China. J Env. Sci 23, 1342–1349.
Zhang, Q., Alfarra, M.R., Worsnop, D.R., Allan, J.D., Coe, H., Canagaratna, M.R.,
Jimenez, J.L., 2005a. Deconvolution and Quantification of Hydrocarbon-like and
Oxygenated Organic Aerosols Based on Aerosol Mass Spectrometry. Environ. Sci.
Technol. 39, 4938–4952.
Zhang, Q., Worsnop, D.R., Canagaratna, M.R., Jimenez, J.L., 2005b. Hydrocarbon-like
and oxygenated organic aerosols in Pittsburgh: insights into sources and processes of
organic aerosols. Atmos. Chem. Phys. 5, 3289–3311.
Zhang, Y.Y., Müller, L., Winterhalter, R., Moortgat, G.K., Hoffmann, T., Pöschl, U., 2010.
Seasonal cycle and temperature dependence of pinene oxidation products,
dicarboxylic acids and nitrophenols in fine and coarse air particulate matter. Atmos.
Chem. Phys. 10, 7859–7873.
References
173
References
174
i
Annexes
Annexes
ii
Annexes
iii
Annex A
Table A. 1. Sampling details of the intensive field campaign II.
Sample Sampling Period Observations
AVE 1 23-11-2009 10:45 30-11-2009 10:45
AVE 2 30-11-2009 11:05 07-12-2009 11:05
AVE 3 07-12-2009 11:26 14-12-2009 11:30
AVE 4 14-12-2009 12:00 21-12-2009 12:00
AVE 5 21-12-2009 12:30 28-12-2009 10:10 Field blank
AVE 6 28-12-2009 10:30 04-01-2010 11:27 Field blank
AVE 7 11-01-2010 10:10 18-01-2010 10:10
AVE 8 18-01-2010 10:45 25-01-2010 10:45 Field blank
AVE 9 26-01-2010 11:15 02-02-2010 11:15
AVE 10 02-02-2010 11:35 09-02-2010 11:35
AVE 11 09-02-2010 11:58 12-02-2010 09:40 Engine failure of the sampler
AVE 12 15-02-2010 10:11 22-02-2010 09:52 Field blank
AVE 13 22-02-2010 10:15 01-03-2010 10:00 Field blank
AVE 14 08-03-2010 14:25 15-03-2010 14:25
AVE 15 15-03-2010 14:44 22-03-2010 10:00 Engine failure of the sampler
AVE 16 30-03-2010 10:00 06-04-2010 10:05
AVE 17 06-04-2010 10:20 13-04-2010 10:18
AVE 18 13-04-2010 10:45 20-04-2010 10:45
AVE 19 20-04-2010 11:05 27-04-2010 11:15
AVE 20 27-04-2010 11:32 04-05-2010 11:26
AVE 21 04-05-2010 11:47 11-05-2010 11:45
AVE 22 11-05-2010 12:13 18-05-2010 12:45
AVE 23 18-05-2010 12:35 25-05-2010 12:45
AVE 24 25-05-2010 13:15 01-06-2010 10:45 Field blank
AVE 25 01-06-2010 11:50 08-06-2010 11:43
AVE 26 08-06-2010 12:09 15-06-2010 10:35 Field blank
AVE 27 15-06-2010 11:40 22-06-2010 10:46
AVE 28 22-06-2010 11:07 29-06-2010 10:47
AVE 29 29-06-2010 11:14 30-06-2010 10:05 Engine failure of the sampler
AVE 30 05-07-2010 10:54 12-07-2010 10:42
AVE 31 12-07-2010 11:00 19-07-2010 11:00
AVE 32 19-07-2010 11:09 26-07-2010 11:00
AVE 33 26-07-2010 11:20 02-08-2010 11:24
AVE 34 02-08-2010 11:45 09-08-2010 11:46
AVE 35 09-08-2010 12:11 16-08-2010 12:18
AVE 36 16-08-2010 12:35 19-08-2010 09:46 Engine failure of the sampler
AVE 37 19-08'2010 12:00 26-08-2010 12:00 Field blank
AVE 38 06-09-2010 11:20 13-09-2010 11:15
AVE 39 13-09-2010 11:36 20-09-2010 11:30
AVE 40 20-09-2010 11:43 27-09-2010 12:05
AVE 41 27-09-2010 12:21 04-10-2010 12:25
AVE 42 04-10-2010 12:47 11-10-2010 12:48
AVE 43 11-10-2010 13:11 18-10-2010 13:15
AVE 44 18-10-2010 13:40 22-10-2010 23:30 Engine failure of the sampler
AVE 45 25-10-2010 09:15 01-11-2010 09:30 Field blank
AVE 46 02-11-2010 10:15 09-11-2010 10:28
AVE 47 09-11-2010 10:45 16-11-2010 11:10
AVE 48 16-11-2010 11:28 23-11-2010 11:45
AVE 49 23-11-2010 12:02 30-11-2010 12:05
AVE 50 30-11-2010 12:25 07-12-2010 12:58
AVE 51 07-12-2010 13:20 13-12-2010 11:30 Engine failure of the sampler
AVE 52 13-12-2010 14:50 20-12-2010 10:25 Field blank
Annexes
iv
Table A. 2. Sampling details of the intensive field campaign II (Cont.).
Sample Sampling Period Observations
AVE 53 20-12-2010 13:10 27-12-2010 10:15
AVE 54 27-12-2010 10:36 03-01-2011 10:50
AVE 55 03-01-2011 11:10 07-01-2011 01:30 Engine failure of the sampler
AVE 56 11-01-2011 10:45 18-01-2011 10:45
AVE 57 18-01-2011 11:00 25-01-2011 10:59
AVE 58 25-01-2011 11:15 01-02-2011 11:25
AVE 59 01-02-2011 11:40 08-02-2011 11:54 Engine failure of the sampler
AVE 60 08-02-2011 12:15 15-02-2011 10:26 Field blank
AVE 61 15-02-2011 11:00 22-02-2011 10:50
AVE 62 22-02-2011 11:10 01-03-2011 11:02
AVE 63 01-03-2011 11:18 08-03-2011 11:45
AVE 64 08-03-2011 12:20 15-03-2011 11:30
AVE 65 15-03-2011 11:45 22-03-2011 12:05
AVE 66 22-03-2011 12:26 29-03-2011 12:08
Annexes
v
Annex B
Fig. B. 1. Air mass backward trajectories ending at Aveiro at distinct altitudes (>500 m a.g.l.) during each week in Summer 2008 (SU08-1 – SU08-4), Spring 2009 (SP09-1 –
SP09-4) and Summer 2009 (SU09-1 – SU09-4).
SU08-1 SU08-2 SU08-3
SU08-4
Annexes
vi
Fig. B. 1. Continued.
SP09-1 SP09-2 SP09-3
SP09-4
Annexes
vii
Fig. B. 1. Continued.
SU09-1 SU09-2 SU09-3
SU09-4
Annexes
viii
Annexes
ix
Annex C
Fig. C. 1. Air mass backward trajectories ending at Aveiro at distinct altitudes (>500 m a.g.l.) during each week in Autumn 2009 (AVE 1 – AVE 4), Winter 2010 (AVE 7 –
AVE 14), Spring 2010 (AVE 16 – AVE 27), Summer 2010 (AVE 28 – AVE 39), Autumn 2010 (AVE 40 – AVE 50), Winter (AVE 53 – AVE 65), and Spring
2011(AVE 66) seasons.
Annexes
x
Fig. C. 1. Continued.
Annexes
xi
Fig. C. 1. Continued.
Annexes
xii
Fig. C. 1. Continued.
Annexes
xiii
Fig. C. 1. Continued.
Annexes
xiv
Fig. C. 1. Continued.
Annexes
xv
Fig. C. 1. Continued.
Annexes
xvi
Fig. C. 1. Continued.
Annexes
xvii
Annex D
Fig. D. 1. EEM spectra (specific fluorescence intensity (g
−1 C L) versus excitation and emission wavelengths (nm)) of the WSOC fractions extracted from the aerosol samples
collected in each week during Autumn 2009 (AVE 1 – AVE 4), Winter 2010 (AVE 7 – AVE 14), Spring 2010 (AVE 16 – AVE 27), Summer 2010 (AVE 28 – AVE
39), Autumn 2010 (AVE 40 – AVE 50), Winter (AVE 53 – AVE 65), and Spring 2011(AVE 66) seasons.
Annexes
xviii
Fig. D. 1. Continued.
Annexes
xix
Fig. D. 1. Continued.
Annexes
xx
Fig. D. 1. Continued.
Annexes
xxi
Fig. D. 1. Continued.
Annexes
xxii
Fig. D. 1. Continued.
Annexes
xxiii
Fig. D. 1. Continued.
Annexes
xxiv
Fig. D. 1. Continued.
Top Related